https://www.hlt.inesc-id.pt/wiki/api.php?action=feedcontributions&user=Ramon&feedformat=atomHLT@INESC-ID - User contributions [en]2024-03-29T15:48:48ZUser contributionsMediaWiki 1.41.0https://www.hlt.inesc-id.pt/wiki/index.php?title=Ramon_Fernandez_Astudillo&diff=7869Ramon Fernandez Astudillo2016-09-28T08:24:08Z<p>Ramon: </p>
<hr />
<div>{{infobox|name=Ramón Fernandez Astudillo<br />
|username=ramon<br />
|contact=ramon.astudillo<br />
|phone=+351-213-100-315<br />
|fax=+351-213-145-843<br />
}}<br />
<!-- YOUR SHORT BIO SHOULD BE HERE --><br />
I obtained the industrial engineering degree with specialization electronics in automatic regulation at the ''Escuela Politecnica Superior de Ingenieria de Gijon'' (Spain) in 2005, completing the last year and final thesis of this degree with an Erasmus scholarship at the ''Technische Universität Berlin''. In 2006 I worked as an intern at Peiker Acustic researching model-based speech enhancement. On this same year I was awarded with a ''La Caixa'' and the German Academic Exchange Service (DAAD) scholarship for research towards the Ph.D. degree at the [https://www.emsp.tu-berlin.de/menue/mitarbeiter/ehemalige_beschaeftigte/ EMSP] department of the ''Technische Universität Berlin''. I obtained the title with distinction in 2010 in the fields of speech processing and robust automatic speech recognition. In my thesis "Integration of Short-Time Fourier Domain Speech Enhancement and Observation Uncertainty Techniques for Robust Automatic Speech Recognition" I formalized the concept of Short-Time Fourier Uncertainty Propagation ([https://github.com/ramon-astudillo/stft_up_tools STFT-UP]) and its relation to speech enhancement techniques such as the Wiener or Ephraim-Malah filters. In 2010 I was awarded with a FCT Post-Doctoral scholarship to join INESC-ID/L2F where I researched both robust speech recognition and robust natural language processing speech applications mixing deep learning and latent variable models. From 2016 I am Senior Research Scientist at [https://unbabel.com/about/ Unbabel], keeping also an affiliation to INESC-ID as associate researcher. I am also co-founder of the Special Interest group for Robust Speech Processing ([https://wiki.inria.fr/rosp/Main_Page RoSP-SIG]) and organizer of the LxMLS summer schools [http://lxmls.it.pt/2015/ 2015], [http://lxmls.it.pt/2016/ 2016]. <br />
<br />
I am an ISCA/ACL member and review in signal processing and machine learning journals and congresses (IEEE-TASLP/SPL, CSL, Interspeech, NIPS and EMNLP among others). It is hard to find things that do not interest me, but artificial and crowd intelligence seem particularly motivating in this moment in history. <br />
<br />
You can find my GitHub repo [https://github.com/ramon-astudillo here]<br />
<br />
<inesc-id what='person' id='1289'></inesc-id><br />
<br />
<br />
== Special Session at Interspeech 2015==<br />
<br />
[[Robust Speech Processing using Observation Uncertainty and Uncertainty Propagation]]<br />
<br />
== Old Publications ==<br />
<br />
A complete list of my publications can be found [http://www.astudillo.com/ramon/research/ here], in particular, my doctoral thesis can be found [http://opus.kobv.de/tuberlin/volltexte/2010/2676/pdf/astudillo_ramon.pdf here]<br />
<br />
== Around the Web ==<br />
My LinkedIn profile can be found [http://www.linkedin.com/in/ramonastudillo here] (I only add people I meet in person).<br />
<br />
[[category:People]]<br />
[[category:Researchers]]</div>Ramonhttps://www.hlt.inesc-id.pt/wiki/index.php?title=Ramon_Fernandez_Astudillo&diff=7658Ramon Fernandez Astudillo2015-03-11T18:41:58Z<p>Ramon: </p>
<hr />
<div>{{infobox|name=Ramón Fernandez Astudillo<br />
|username=ramon<br />
|contact=ramon.astudillo<br />
|phone=+351-213-100-315<br />
|fax=+351-213-145-843<br />
}}<br />
<!-- YOUR SHORT BIO SHOULD BE HERE --><br />
I obtained the industrial engineering degree with specialization electronics in automatic regulation at the ''Escuela Politecnica Superior de Ingenieria de Gijon'' (Spain) in 2005, completing the last two years of this degree with an Erasmus scholarship at the ''Technische Universität Berlin''. In 2006 I worked as an intern at Peiker Acustic researching model-based speech enhancement. On this same year I was awarded with a ''La Caixa'' and the German Academic Exchange Service (DAAD) scholarship for research towards the Ph.D. degree. I obtained the title with distinction from the ''Technische Universität Berlin'' in 2010 in the fields of speech processing and robust automatic speech recognition. In my thesis "Integration of Short-Time Fourier Domain Speech Enhancement and Observation Uncertainty Techniques for Robust Automatic Speech Recognition" I formalized the concept of short-time Fourier uncertainty propagation and introduced various closed from solutions for it. I am currently a Post.-Doc. researcher here at INESC-ID/L2F researching both on robust speech recognition and robust natural language processing speech applications in a Bayesian setting. I am an ISCA member and reviewer of IEEE-TASLP/SPL as well as CSL. My interests span statistics, signal processing, machine learning and in general both human and machine perception of the environment.<br />
<br />
You can find my GitHub repo [https://github.com/ramon-astudillo here]<br />
<br />
<inesc-id what='person' id='1289'></inesc-id><br />
<br />
<br />
== Special Session at Interspeech 2015==<br />
<br />
[[Robust Speech Processing using Observation Uncertainty and Uncertainty Propagation]]<br />
<br />
== Old Publications ==<br />
<br />
A complete list of my publications can be found [http://www.astudillo.com/ramon/research/ here], in particular, my doctoral thesis can be found [http://opus.kobv.de/tuberlin/volltexte/2010/2676/pdf/astudillo_ramon.pdf here]<br />
<br />
== Around the Web ==<br />
My LinkedIn profile can be found [http://www.linkedin.com/in/ramonastudillo here] (I only add people I meet in person).<br />
<br />
[[category:People]]<br />
[[category:Researchers]]</div>Ramonhttps://www.hlt.inesc-id.pt/wiki/index.php?title=Opportunities_(all_announcements)&diff=7657Opportunities (all announcements)2015-03-11T18:37:20Z<p>Ramon: /* Post-Doc Fellowship in Project INSIDE */</p>
<hr />
<div>__NOTOC__<br />
== Thinking about Joining Us? ==<br />
INESC ID is one of the largest and most dynamic research institutes in Portugal in the areas of telecommunications, information systems and electronics with the status of associated laboratory given by the Portuguese government. The ''Laboratório de sistemas de Língua Falada'' (L2F), or Spoken Language Systems Lab provides an interdisciplinary and creative work environment active in multiple areas of speech and language research. These include <br />
<br />
* speech recognition, synthesis, enhancement and understanding, speech to speech translation<br />
<br />
* speaker, language and accent identification, computational paralinguistics<br />
<br />
* music genre classification, audio finger-printing and retrieval, multi-document summarization of multimedia (audio + video + text)<br />
<br />
* machine translation, summarization, sentiment analysis<br />
<br />
* natural language generation, syntactic and semantic processing, named entity retrieval<br />
<br />
* multimodal dialogue systems, question answering, etc. <br />
<br />
Have a look at our recent [https://www.l2f.inesc-id.pt/wiki/index.php/Projects projects] and [https://www.l2f.inesc-id.pt/wiki/index.php/Demos demos]. <br />
<br />
L2F has its own powerful IT infrastructure within INESC-ID including <br />
<br />
* Robust and fast shared filesystems for storage and processing of very large data-sets<br />
<br />
* A computer cluster with Hadoop and condor among other parallel computing utilities <br />
<br />
* Dedicated computing units with GPUs and intel-phi CPUs<br />
<br />
* VPN, ssh as well as hosting for CVS, SVN and Git repositories.<br />
<br />
This infrastructure also provides online services to various academic and private sector partners. We have a strong cooperation with our spin-off [http://www.voiceinteraction.pt/ VoiceInteraction] and with well established industrial partners such as [http://www.microsoft.com/pt-pt/mldc/ Microsoft], as well as startups from our own group such as [https://unbabel.com/ Unbabel]. <br />
<br />
Finally, it is also worth mentioning that Lisbon offers a high quality of life while being one of the least expensive capitals in Europe, test it yourself [http://www.expatistan.com/cost-of-living here]. <br />
<br />
Want to joint us?. Have a look at the opportunities below. If there aren't any at the time, feel also free to send us an email to find out about potential openings or partnerships. <br />
<br />
<!-- Ramon: I commented this because it did not update --><br />
<!-- <inesc-id what='calls' cc='IP02'></inesc-id> --><br />
<!-- <inesc-id what='calls' cc='IP02' showclosed='yes'></inesc-id>--><br />
<!-- PLEASE POST NEW POSITION ANNOUNCEMENTS THROUGH www.inesc-id.pt --><br />
<br />
== Ph.D. Program CMU-Portugal ==<br />
<br />
* [[Ph.D. Program Carnegie Mellon - PORTUGAL in the area of Language and Information Technologies|'''Ph.D. Program Carnegie Mellon-Portugal''' in the area of Language and Information Technologies]]<br />
<br />
== Post-Doc Fellowship in Project INSIDE ==<br />
<br />
This is a 3 year position, starting in April 2015, earliest.<br />
<br />
INSIDE is a joint research project involving Portuguese academic/industrial partners and the Carnegie Mellon University. The project explores symbiotic child-robot interactions in joint cooperative activities, with application to the treatment of children with Autism Spectrum Disorder impairments.<br />
<br />
Your role would be related to speech processing in environmentally challenging conditions and with limited adaptation data available. Research will be focused on static and dynamic adaptation of models starting from our current research on deep learning and observation uncertainty techniques.<br />
<br />
We are looking for candidates with following experience:<br />
<br />
* Robust ASR/KWS<br />
* Neural networks<br />
* Python, bash scripting and C/C++<br />
<br />
Additional experience welcome<br />
<br />
* Computational Paralinguistics<br />
* Python-theano<br />
* Knowledge of the Portuguese language<br />
* Missing features / Observation uncertainty<br />
<br />
If interested, send a curriculum vitae (highlighting publications relevant to the position) and a motivation letter to ramon.astudillo@l2f.inesc-id.pt before March 27.<br />
<br />
== Opportunities for Students ==<br />
<br />
'''[[Master Theses Themes at L2F]]''' (2013/14)<br><br><br />
<!--<br />
<br> '''New courses of Portuguese as Second Language - starting 2nd October 2012''' - by Ciberescola with the activities in colaboration with INESC-ID. Visit: http://www.cibercursoslp.com/inscricoes.html <br />
--><br />
<br />
== Other Information ==<br />
<br />
* A [[Positions|complete list]] of calls posted on [http://www.inesc-id.pt/intranet/bolsas/list_fellowships.php INESC ID's website], with both open and closed calls.<br />
* [[Older Calls]]</div>Ramonhttps://www.hlt.inesc-id.pt/wiki/index.php?title=Opportunities_(all_announcements)&diff=7656Opportunities (all announcements)2015-03-11T18:35:49Z<p>Ramon: /* Post-Doc Fellowship in Project INSIDE */</p>
<hr />
<div>__NOTOC__<br />
== Thinking about Joining Us? ==<br />
INESC ID is one of the largest and most dynamic research institutes in Portugal in the areas of telecommunications, information systems and electronics with the status of associated laboratory given by the Portuguese government. The ''Laboratório de sistemas de Língua Falada'' (L2F), or Spoken Language Systems Lab provides an interdisciplinary and creative work environment active in multiple areas of speech and language research. These include <br />
<br />
* speech recognition, synthesis, enhancement and understanding, speech to speech translation<br />
<br />
* speaker, language and accent identification, computational paralinguistics<br />
<br />
* music genre classification, audio finger-printing and retrieval, multi-document summarization of multimedia (audio + video + text)<br />
<br />
* machine translation, summarization, sentiment analysis<br />
<br />
* natural language generation, syntactic and semantic processing, named entity retrieval<br />
<br />
* multimodal dialogue systems, question answering, etc. <br />
<br />
Have a look at our recent [https://www.l2f.inesc-id.pt/wiki/index.php/Projects projects] and [https://www.l2f.inesc-id.pt/wiki/index.php/Demos demos]. <br />
<br />
L2F has its own powerful IT infrastructure within INESC-ID including <br />
<br />
* Robust and fast shared filesystems for storage and processing of very large data-sets<br />
<br />
* A computer cluster with Hadoop and condor among other parallel computing utilities <br />
<br />
* Dedicated computing units with GPUs and intel-phi CPUs<br />
<br />
* VPN, ssh as well as hosting for CVS, SVN and Git repositories.<br />
<br />
This infrastructure also provides online services to various academic and private sector partners. We have a strong cooperation with our spin-off [http://www.voiceinteraction.pt/ VoiceInteraction] and with well established industrial partners such as [http://www.microsoft.com/pt-pt/mldc/ Microsoft], as well as startups from our own group such as [https://unbabel.com/ Unbabel]. <br />
<br />
Finally, it is also worth mentioning that Lisbon offers a high quality of life while being one of the least expensive capitals in Europe, test it yourself [http://www.expatistan.com/cost-of-living here]. <br />
<br />
Want to joint us?. Have a look at the opportunities below. If there aren't any at the time, feel also free to send us an email to find out about potential openings or partnerships. <br />
<br />
<!-- Ramon: I commented this because it did not update --><br />
<!-- <inesc-id what='calls' cc='IP02'></inesc-id> --><br />
<!-- <inesc-id what='calls' cc='IP02' showclosed='yes'></inesc-id>--><br />
<!-- PLEASE POST NEW POSITION ANNOUNCEMENTS THROUGH www.inesc-id.pt --><br />
<br />
== Ph.D. Program CMU-Portugal ==<br />
<br />
* [[Ph.D. Program Carnegie Mellon - PORTUGAL in the area of Language and Information Technologies|'''Ph.D. Program Carnegie Mellon-Portugal''' in the area of Language and Information Technologies]]<br />
<br />
== Post-Doc Fellowship in Project INSIDE ==<br />
<br />
This is a 3 year position, starting in April 2015.<br />
<br />
INSIDE is a joint research project involving Portuguese academic/industrial partners and the Carnegie Mellon University. The project explores symbiotic child-robot interactions in joint cooperative activities, with application to the treatment of children with Autism Spectrum Disorder impairments.<br />
<br />
Your role would be related to speech processing in environmentally challenging conditions and with limited adaptation data available. Research will be focused on static and dynamic adaptation of models starting from our current research on deep learning and observation uncertainty techniques.<br />
<br />
We are looking for candidates with following experience:<br />
<br />
* Robust ASR/KWS<br />
* Neural networks<br />
* Python, bash scripting and C/C++<br />
<br />
Additional experience welcome<br />
<br />
* Computational Paralinguistics<br />
* Python-theano<br />
* Knowledge of the Portuguese language<br />
* Missing features / Observation uncertainty<br />
<br />
If interested, send a curriculum vitae (highlighting publications relevant to the position) and a motivation letter to ramon.astudillo@l2f.inesc-id.pt before March 27.<br />
<br />
== Opportunities for Students ==<br />
<br />
'''[[Master Theses Themes at L2F]]''' (2013/14)<br><br><br />
<!--<br />
<br> '''New courses of Portuguese as Second Language - starting 2nd October 2012''' - by Ciberescola with the activities in colaboration with INESC-ID. Visit: http://www.cibercursoslp.com/inscricoes.html <br />
--><br />
<br />
== Other Information ==<br />
<br />
* A [[Positions|complete list]] of calls posted on [http://www.inesc-id.pt/intranet/bolsas/list_fellowships.php INESC ID's website], with both open and closed calls.<br />
* [[Older Calls]]</div>Ramonhttps://www.hlt.inesc-id.pt/wiki/index.php?title=Robust_Speech_Processing_using_Observation_Uncertainty_and_Uncertainty_Propagation&diff=7632Robust Speech Processing using Observation Uncertainty and Uncertainty Propagation2015-03-09T19:56:38Z<p>Ramon: /* Call for Papers for a Special Session at Interspeech 2015 */</p>
<hr />
<div>== Call for Papers for a Special Session at Interspeech 2015 ==<br />
<br />
<br />
<br />
We are pleased to announce a special session at [http://interspeech2015.org/ Interspeech 2015] in Dresden focusing on Observation Uncertainty (OU) and Uncertainty Propagation (UP) research in speech processing with<br />
<br />
'''*Deadline March 20*'''<br />
<br />
Observation Uncertainty (OU) and Uncertainty Propagation (UP) techniques for robust speech processing have shown notable success across various sub-domains as e.g. ASR or Speaker Recognition. Despite its healthy status, OU/UP research in speech processing is currently fragmented across various disciplines and application domains. Consequently, research papers on uncertainty have to compete with other techniques in e.g. ASR or Speaker Identification tracks. This generally discourages the proposal of novel or bold ideas pertaining to the topic. This is particularly relevant in the current environment in which the strong technology transfer and real-data-rich problems force disciplines to adapt themselves more quickly. In particular, although investigations on how to port OU/UP techniques to Deep Neural Networks (DNNs) exist, this area is still yet to be studied. Given the disruptive character of DNNs and the fact that they are not amenable to classic adaptation techniques, this poses a very interesting application dimension for OU/UP techniques. For these reasons, we have proposed a special session devoted to uncertainty as an opportunity for researchers of different sub-domains of speech processing to share ideas, explore new application domains as e.g. paralinguistics, or text-based uncertainties and consolidate OU/UP research in DNNs. <br />
<br />
The session would cover <br />
<br />
* Methods exploiting observation and parameter uncertainty as well as uncertainty propagation for robust speech processing e.g., for inference, learning, adaptation or model selection.<br />
<br />
* We specially welcome papers on the topic deep learning and uncertainty. We are confident this is the right moment to tackle this issue and already expect works on the topic.<br />
<br />
* Uncertainty in training data: “BIG” data is upon us, but the quality of that ‘found’ data is not the same as carefully collected data. What are the best ways to address data uncertainty when ground-truth is unknown? <br />
<br />
* All sub-fields are welcome, since this is not restricted to only ASR and speaker recognition communities but also works in e.g., paralinguistics or transcription level uncertainties (lattices).<br />
<br />
We would also like to use this opportunity to motivate the sharing of software tools based on uncertainty techniques among the community as well as reproducible research. For this purpose we will provide software for uncertainty-related techniques, such as tools for the estimation of observation uncertainties and for their application to ASR, through the wiki page of the Robust Speech Processing Special Interest Group of ISCA (RoSP-SIG).<br />
<br />
https://wiki.inria.fr/rosp/Software<br />
<br />
All updates on the session progress and available tools will be reported on this website.<br />
<br />
'''Update 2015-02-10''': Tools for uncertainty propagation and observation uncertainty using HTK/Kaldi released (see WiKi link above). <br />
<br />
Do not hesitate to contact us if you have any question: <br />
<br />
Ramón Fernandez Astudillo (ramon.astudillo@inesc-id.pt)<br />
<br />
Shinji Watanabe (watanabe@merl.com)<br />
<br />
Ahmed Hussen AbdelAziz (Ahmed.HussenAbdelAziz@rub.de)<br />
<br />
Dorothea Kolossa (dorothea.kolossa@rub.de)</div>Ramonhttps://www.hlt.inesc-id.pt/wiki/index.php?title=Robust_Speech_Processing_using_Observation_Uncertainty_and_Uncertainty_Propagation&diff=7631Robust Speech Processing using Observation Uncertainty and Uncertainty Propagation2015-03-09T19:56:19Z<p>Ramon: /* Call for Papers for a Special Session at Interspeech 2015 */</p>
<hr />
<div>== Call for Papers for a Special Session at Interspeech 2015 ==<br />
<br />
<br />
<br />
We are pleased to announce a special session at [http://interspeech2015.org/ Interspeech 2015] in Dresden focusing on Observation Uncertainty (OU) and Uncertainty Propagation (UP) research in speech processing with<br />
<br />
'''*Deadline March 20*'''<br />
<br />
Observation Uncertainty (OU) and Uncertainty Propagation (UP) techniques for robust speech processing have shown notable success across various sub-domains as e.g. ASR or Speaker Recognition. Despite its healthy status, OU/UP research in speech processing is currently fragmented across various disciplines and application domains. Consequently, research papers on uncertainty have to compete with other techniques in e.g. ASR or Speaker Identification tracks. This generally discourages the proposal of novel or bold ideas pertaining to the topic. This is particularly relevant in the current environment in which the strong technology transfer and real-data-rich problems force disciplines to adapt themselves more quickly. In particular, although investigations on how to port OU/UP techniques to Deep Neural Networks (DNNs) exist, this area is still yet to be studied. Given the disruptive character of DNNs and the fact that they are not amenable to classic adaptation techniques, this poses a very interesting application dimension for OU/UP techniques. For these reasons, we have proposed a special session devoted to uncertainty as an opportunity for researchers of different sub-domains of speech processing to share ideas, explore new application domains as e.g. paralinguistics, or text-based uncertainties and consolidate OU/UP research in DNNs. <br />
<br />
The session would cover <br />
<br />
* Methods exploiting observation and parameter uncertainty as well as uncertainty propagation for robust speech processing e.g., for inference, learning, adaptation or model selection.<br />
<br />
* We specially welcome papers on the topic deep learning and uncertainty. We are confident this is the right moment to tackle this issue and already expect works on the topic.<br />
<br />
* Uncertainty in training data: “BIG” data is upon us, but the quality of that ‘found’ data is not the same as carefully collected data. What are the best ways to address data uncertainty when ground-truth is unknown? <br />
<br />
* All sub-fields are welcome, since this is not restricted to only ASR and speaker recognition communities but also works in e.g., paralinguistics or transcription level uncertainties (lattices).<br />
<br />
We would also like to use this opportunity to motivate the sharing of software tools based on uncertainty techniques among the community as well as reproducible research. For this purpose we will provide software for uncertainty-related techniques, such as tools for the estimation of observation uncertainties and for their application to ASR, through the wiki page of the Robust Speech Processing Special Interest Group of ISCA (RoSP-SIG).<br />
<br />
https://wiki.inria.fr/rosp/Software<br />
<br />
All updates on the session progress and available tools will be reported on this website.<br />
<br />
''Update 2015-02-10'': Tools for uncertainty propagation and observation uncertainty using HTK/Kaldi released (see WiKi link above). <br />
<br />
Do not hesitate to contact us if you have any question: <br />
<br />
Ramón Fernandez Astudillo (ramon.astudillo@inesc-id.pt)<br />
<br />
Shinji Watanabe (watanabe@merl.com)<br />
<br />
Ahmed Hussen AbdelAziz (Ahmed.HussenAbdelAziz@rub.de)<br />
<br />
Dorothea Kolossa (dorothea.kolossa@rub.de)</div>Ramonhttps://www.hlt.inesc-id.pt/wiki/index.php?title=Robust_Speech_Processing_using_Observation_Uncertainty_and_Uncertainty_Propagation&diff=7551Robust Speech Processing using Observation Uncertainty and Uncertainty Propagation2015-03-02T12:18:17Z<p>Ramon: /* Call for Papers for a Special Session at Interspeech 2015 */</p>
<hr />
<div>== Call for Papers for a Special Session at Interspeech 2015 ==<br />
<br />
<br />
<br />
We are pleased to announce a special session at [http://interspeech2015.org/ Interspeech 2015] in Dresden focusing on Observation Uncertainty (OU) and Uncertainty Propagation (UP) research in speech processing with<br />
<br />
'''*Deadline March 20*'''<br />
<br />
Observation Uncertainty (OU) and Uncertainty Propagation (UP) techniques for robust speech processing have shown notable success across various sub-domains as e.g. ASR or Speaker Recognition. Despite its healthy status, OU/UP research in speech processing is currently fragmented across various disciplines and application domains. Consequently, research papers on uncertainty have to compete with other techniques in e.g. ASR or Speaker Identification tracks. This generally discourages the proposal of novel or bold ideas pertaining to the topic. This is particularly relevant in the current environment in which the strong technology transfer and real-data-rich problems force disciplines to adapt themselves more quickly. In particular, although investigations on how to port OU/UP techniques to Deep Neural Networks (DNNs) exist, this area is still yet to be studied. Given the disruptive character of DNNs and the fact that they are not amenable to classic adaptation techniques, this poses a very interesting application dimension for OU/UP techniques. For these reasons, we have proposed a special session devoted to uncertainty as an opportunity for researchers of different sub-domains of speech processing to share ideas, explore new application domains as e.g. paralinguistics, or text-based uncertainties and consolidate OU/UP research in DNNs. <br />
<br />
The session would cover <br />
<br />
* Methods exploiting observation and parameter uncertainty as well as uncertainty propagation for robust speech processing e.g., for inference, learning or adaptation<br />
<br />
* We specially welcome papers on the topic deep learning and uncertainty. We are confident this is the right moment to tackle this issue and already expect works on the topic<br />
<br />
* Uncertainty in training data: “BIG” data is upon us, but the quality of that ‘found’ data is not the same as carefully collected data. What are the best ways to address data uncertainty when ground-truth is unknown? <br />
<br />
* All sub-fields are welcome, since this is not restricted to only ASR and speaker recognition communities but also works in e.g., paralinguistics or transcription level uncertainties (lattices).<br />
<br />
We would also like to use this opportunity to motivate the sharing of software tools based on uncertainty techniques among the community as well as reproducible research. For this purpose we will provide software for uncertainty-related techniques, such as tools for the estimation of observation uncertainties and for their application to ASR, through the wiki page of the Robust Speech Processing Special Interest Group of ISCA (RoSP-SIG).<br />
<br />
https://wiki.inria.fr/rosp/Software<br />
<br />
All updates on the session progress and available tools will be reported on this website.<br />
<br />
Update 2015-02-10: Tools for uncertainty propagation and observation uncertainty using HTK/Kaldi released (see WiKi link above). <br />
<br />
Do not hesitate to contact us if you have any question: <br />
<br />
Ramón Fernandez Astudillo (ramon.astudillo@inesc-id.pt)<br />
<br />
Shinji Watanabe (watanabe@merl.com)<br />
<br />
Ahmed Hussen AbdelAziz (Ahmed.HussenAbdelAziz@rub.de)<br />
<br />
Dorothea Kolossa (dorothea.kolossa@rub.de)</div>Ramonhttps://www.hlt.inesc-id.pt/wiki/index.php?title=Robust_Speech_Processing_using_Observation_Uncertainty_and_Uncertainty_Propagation&diff=7550Robust Speech Processing using Observation Uncertainty and Uncertainty Propagation2015-03-02T12:17:29Z<p>Ramon: /* Call for Papers for a Special Session at Interspeech 2015 *Deadline March 20* */</p>
<hr />
<div>== Call for Papers for a Special Session at Interspeech 2015 ==<br />
<br />
'''*Deadline March 20*'''<br />
<br />
We are pleased to announce a special session at [http://interspeech2015.org/ Interspeech 2015] in Dresden focusing on Observation Uncertainty (OU) and Uncertainty Propagation (UP) research in speech processing.<br />
<br />
Observation Uncertainty (OU) and Uncertainty Propagation (UP) techniques for robust speech processing have shown notable success across various sub-domains as e.g. ASR or Speaker Recognition. Despite its healthy status, OU/UP research in speech processing is currently fragmented across various disciplines and application domains. Consequently, research papers on uncertainty have to compete with other techniques in e.g. ASR or Speaker Identification tracks. This generally discourages the proposal of novel or bold ideas pertaining to the topic. This is particularly relevant in the current environment in which the strong technology transfer and real-data-rich problems force disciplines to adapt themselves more quickly. In particular, although investigations on how to port OU/UP techniques to Deep Neural Networks (DNNs) exist, this area is still yet to be studied. Given the disruptive character of DNNs and the fact that they are not amenable to classic adaptation techniques, this poses a very interesting application dimension for OU/UP techniques. For these reasons, we have proposed a special session devoted to uncertainty as an opportunity for researchers of different sub-domains of speech processing to share ideas, explore new application domains as e.g. paralinguistics, or text-based uncertainties and consolidate OU/UP research in DNNs. <br />
<br />
The session would cover <br />
<br />
* Methods exploiting observation and parameter uncertainty as well as uncertainty propagation for robust speech processing e.g., for inference, learning or adaptation<br />
<br />
* We specially welcome papers on the topic deep learning and uncertainty. We are confident this is the right moment to tackle this issue and already expect works on the topic<br />
<br />
* Uncertainty in training data: “BIG” data is upon us, but the quality of that ‘found’ data is not the same as carefully collected data. What are the best ways to address data uncertainty when ground-truth is unknown? <br />
<br />
* All sub-fields are welcome, since this is not restricted to only ASR and speaker recognition communities but also works in e.g., paralinguistics or transcription level uncertainties (lattices).<br />
<br />
We would also like to use this opportunity to motivate the sharing of software tools based on uncertainty techniques among the community as well as reproducible research. For this purpose we will provide software for uncertainty-related techniques, such as tools for the estimation of observation uncertainties and for their application to ASR, through the wiki page of the Robust Speech Processing Special Interest Group of ISCA (RoSP-SIG).<br />
<br />
https://wiki.inria.fr/rosp/Software<br />
<br />
All updates on the session progress and available tools will be reported on this website.<br />
<br />
Update 2015-02-10: Tools for uncertainty propagation and observation uncertainty using HTK/Kaldi released (see WiKi link above). <br />
<br />
Do not hesitate to contact us if you have any question: <br />
<br />
Ramón Fernandez Astudillo (ramon.astudillo@inesc-id.pt)<br />
<br />
Shinji Watanabe (watanabe@merl.com)<br />
<br />
Ahmed Hussen AbdelAziz (Ahmed.HussenAbdelAziz@rub.de)<br />
<br />
Dorothea Kolossa (dorothea.kolossa@rub.de)</div>Ramonhttps://www.hlt.inesc-id.pt/wiki/index.php?title=Opportunities_(all_announcements)&diff=7527Opportunities (all announcements)2015-02-23T21:10:18Z<p>Ramon: </p>
<hr />
<div>__NOTOC__<br />
== Thinking about Joining Us? ==<br />
INESC ID is one of the largest and most dynamic research institutes in Portugal in the areas of telecommunications, information systems and electronics with the status of associated laboratory given by the Portuguese government. The ''Laboratorio de Lengua Falada'' (L2F), or Spoken Language Systems Lab in English, provides an interdisciplinary and creative work environment active in multiple areas of speech and language research. This include <br />
<br />
* speech enhancement, recognition, synthesis and understanding, speech to speech translation<br />
<br />
* speaker, language and dialect identification, computational paralinguistics<br />
<br />
* music genre classification, audio finger-printing and retrieval, multi-document summarization of multimedia (audio + video + text)<br />
<br />
* machine translation, summarization, sentiment analysis<br />
<br />
* natural language generation, syntactic and semantic processing, named entity retrieval<br />
<br />
* multimodal dialogue systems, question answering, <br />
<br />
Have a look at our recent [https://www.l2f.inesc-id.pt/wiki/index.php/Projects projects] and [https://www.l2f.inesc-id.pt/wiki/index.php/Demos demos]. <br />
<br />
L2F has its own powerful IT infrastructure within INESC-ID including <br />
<br />
* Robust and fast shared filesystems for storage and processing of very large data-sets<br />
<br />
* A computer cluster with Hadoop and condor among other parallel computing utilities <br />
<br />
* Dedicated computing units with GPUs and intel-phi CPUs<br />
<br />
* VPN, ssh as well as hosting for CVS, SVN and Git repositories.<br />
<br />
This infrastructure also provides online services to various academic and private sector partners. We also share resources with our spin-off [http://www.voiceinteraction.pt/ VoiceInteraction] and cooperate with well established industrial partners such as [http://www.microsoft.com/pt-pt/mldc/ Microsoft] as well as startups such as [https://unbabel.com/ Unbabel]. <br />
<br />
Finally, it is also worth mentioning that Lisbon offers a high quality of life while being one of the least expensive capitals in Europe, test it yourself [http://www.expatistan.com/cost-of-living here]. <br />
<br />
Want to joint us?. Have a look at the opportunities below. If there aren't any at the time, feel also free to send us an email to find out about potential openings or partnerships. <br />
<br />
<!-- Ramon: I commented this because it did not update --><br />
<!-- <inesc-id what='calls' cc='IP02'></inesc-id> --><br />
<!-- <inesc-id what='calls' cc='IP02' showclosed='yes'></inesc-id>--><br />
<!-- PLEASE POST NEW POSITION ANNOUNCEMENTS THROUGH www.inesc-id.pt --><br />
<br />
== Ph.D. Program CMU-Portugal ==<br />
<br />
* [[Ph.D. Program Carnegie Mellon - PORTUGAL in the area of Language and Information Technologies|'''Ph.D. Program Carnegie Mellon-Portugal''' in the area of Language and Information Technologies]]<br />
<br />
== Post-Doc Fellowship in Project INSIDE ==<br />
<br />
This is a 3 year position, starting in April 2015.<br />
<br />
INSIDE is a joint research project involving Portuguese academic/industrial partners and the Carnegie Mellon University. The project explores symbiotic child-robot interactions in joint cooperative activities, with application to the treatment of children with Autism Spectrum Disorder impairments.<br />
<br />
Your role would be related to speech processing in environmentally challenging conditions and with limited adaptation data available. Research will be focused on static and dynamic adaptation of models starting from our current research on deep learning and observation uncertainty techniques.<br />
<br />
We are looking for candidates with following experience:<br />
<br />
* Robust ASR/KWS<br />
* Neural networks<br />
* Python, bash scripting and C/C++<br />
<br />
Additional experience welcome<br />
<br />
* Computational Paralinguistics<br />
* Python-theano<br />
* Knowledge of the Portuguese language<br />
* Missing features / Observation uncertainty<br />
<br />
If interested, send a curriculum vitae (highlighting publications relevant to the position) and a motivation letter to ramon.astudillo@l2f.inesc-id.pt before March 15.<br />
<br />
<br />
<br />
== Opportunities for Final Year Students ==<br />
<br />
(currently closed)<br />
<br />
== Other Information ==<br />
<br />
* A [[Positions|complete list]] of calls posted on [http://www.inesc-id.pt/intranet/bolsas/list_fellowships.php INESC ID's website], with both open and closed calls.<br />
* [[Older Calls]]</div>Ramonhttps://www.hlt.inesc-id.pt/wiki/index.php?title=Opportunities_(all_announcements)&diff=7526Opportunities (all announcements)2015-02-23T21:09:23Z<p>Ramon: </p>
<hr />
<div>__NOTOC__<br />
== Thinking about Joining Us? ==<br />
INESC ID is one of the largest and most dynamic research institutes in Portugal in the areas of telecommunications, information systems and electronics with the status of associated laboratory given by the Portuguese government. The ''Laboratorio de Lengua Falada'' (L2F), or Spoken Language Systems Lab in English, provides an interdisciplinary and creative work environment active in multiple areas of speech and language research. This include <br />
<br />
* speech enhancement, recognition, synthesis and understanding, speech to speech translation<br />
<br />
* speaker, language and dialect identification, computational paralinguistics<br />
<br />
* music genre classification, audio finger-printing and retrieval, multi-document summarization of multimedia (audio + video + text)<br />
<br />
* machine translation, summarization, sentiment analysis<br />
<br />
* natural language generation, syntactic and semantic processing, named entity retrieval<br />
<br />
* multimodal dialogue systems, question answering, <br />
<br />
Have a look at our recent [https://www.l2f.inesc-id.pt/wiki/index.php/Projects projects] and [https://www.l2f.inesc-id.pt/wiki/index.php/Demos demos]. <br />
<br />
L2F has its own powerful IT infrastructure within INESC-ID including <br />
<br />
* Robust and fast shared filesystems for storage and processing of very large data-sets<br />
<br />
* A computer cluster with Hadoop and condor among other parallel computing utilities <br />
<br />
* Dedicated computing units with GPUs and intel-phi CPUs<br />
<br />
* VPN, ssh as well as hosting for CVS, SVN and Git repositories.<br />
<br />
This infrastructure also provides online services to various academic and private sector partners. We also share resources with our spin-off [http://www.voiceinteraction.pt/ VoiceInteraction] and cooperate with well established industrial partners such as [http://www.microsoft.com/pt-pt/mldc/ Microsoft] as well as startups such as [https://unbabel.com/ Unbabel]. <br />
<br />
Finally, it is also worth mentioning that Lisbon offers a high quality of life while being one of the least expensive capitals in Europe, test it yourself [http://www.expatistan.com/cost-of-living here]. <br />
<br />
Want to joint us?. Have a look at the opportunities below. If there aren't any at the time, feel also free to send us an email to find out about potential openings or partnerships. <br />
<br />
<!-- Ramon: I commented this because it did not update --><br />
<!-- <inesc-id what='calls' cc='IP02'></inesc-id> --><br />
<!-- <inesc-id what='calls' cc='IP02' showclosed='yes'></inesc-id>--><br />
<!-- PLEASE POST NEW POSITION ANNOUNCEMENTS THROUGH www.inesc-id.pt --><br />
<br />
== Ph.D. Program CMU-Portugal ==<br />
<br />
* [[Ph.D. Program Carnegie Mellon - PORTUGAL in the area of Language and Information Technologies|'''Ph.D. Program Carnegie Mellon-Portugal''' in the area of Language and Information Technologies]]<br />
<br />
== Post-Doc Fellowship in Project INSIDE ==<br />
<br />
This is a 3 year position, starting in April 2015.<br />
<br />
INSIDE is a joint research project involving Portuguese academic/industrial partners and Carnegie Mellon University. The project explores symbiotic child-robot interactions in joint cooperative activities, with application to the treatment of children with Autism Spectrum Disorder impairments.<br />
<br />
Your role would be related to speech processing in environmentally challenging conditions and with limited adaptation data available. Research will be focused on static and dynamic adaptation of models starting from our current research on deep learning and observation uncertainty techniques.<br />
<br />
We are looking for candidates with following experience:<br />
<br />
* Robust ASR/KWS<br />
* Neural networks<br />
* Python, bash scripting and C/C++<br />
<br />
Additional experience welcome<br />
<br />
* Computational Paralinguistics<br />
* Python-theano<br />
* Knowledge of the Portuguese language<br />
* Missing features / Observation uncertainty<br />
<br />
If interested, send a curriculum vitae (highlighting publications relevant to the position) and a motivation letter to ramon.astudillo@l2f.inesc-id.pt before March 15.<br />
<br />
<br />
<br />
== Opportunities for Final Year Students ==<br />
<br />
(currently closed)<br />
<br />
== Other Information ==<br />
<br />
* A [[Positions|complete list]] of calls posted on [http://www.inesc-id.pt/intranet/bolsas/list_fellowships.php INESC ID's website], with both open and closed calls.<br />
* [[Older Calls]]</div>Ramonhttps://www.hlt.inesc-id.pt/wiki/index.php?title=Opportunities_(all_announcements)&diff=7525Opportunities (all announcements)2015-02-23T20:31:12Z<p>Ramon: </p>
<hr />
<div>__NOTOC__<br />
== Thinking about Joining Us? ==<br />
INESC ID is one of the largest and most dynamic research institutes in Portugal in the areas of telecommunications, information systems and electronics with the status of associated laboratory given by the Portuguese government. The Laboratorio de Lengua Falada (L2F), or Spoken Language Systems Lab in English, provides an interdisciplinary and creative work environment active in multiple areas of speech and language research. This include <br />
<br />
* speech enhancement, recognition, synthesis and understanding, speech to speech translation<br />
<br />
* speaker, language and dialect identification, computational paralinguistics<br />
<br />
* music genre classification, audio finger-printing and retrieval, multi-document summarization of multimedia (audio + video + text)<br />
<br />
* machine translation, summarization, sentiment analysis<br />
<br />
* natural language generation, syntactic and semantic processing, named entity retrieval<br />
<br />
* multimodal dialogue systems, question answering, <br />
<br />
Have a look at our recent [https://www.l2f.inesc-id.pt/wiki/index.php/Projects projects] and [https://www.l2f.inesc-id.pt/wiki/index.php/Demos demos]. <br />
<br />
L2F has its own powerful IT infrastructure within INESC-ID including <br />
<br />
* Robust and fast shared filesystems for storage and processing of very large data-sets<br />
<br />
* A computer cluster with Hadoop and condor among other parallel computing utilities <br />
<br />
* Dedicated computing units with GPUs and intel-phi CPUs<br />
<br />
* VPN, ssh as well as hosting for CVS, SVN and Git repositories.<br />
<br />
This infrastructure also provides online services to various academic and private sector partners. We also share resources with our spin-off [http://www.voiceinteraction.pt/ VoiceInteraction] and cooperate with well established industrial partners such as [http://www.microsoft.com/pt-pt/mldc/ Microsoft] as well as startups such as [https://unbabel.com/ Unbabel]. <br />
<br />
Finally, it is also worth mentioning that Lisbon offers a high quality of life while being one of the least expensive capitals in Europe, test it yourself [http://www.expatistan.com/cost-of-living here]. <br />
<br />
Want to joint us?. Have a look at the opportunities below. If there aren't any at the time, feel also free to send us an email to find out about potential openings or partnerships. <br />
<br />
<!-- Ramon: I commented this because it did not update --><br />
<!-- <inesc-id what='calls' cc='IP02'></inesc-id> --><br />
<!-- <inesc-id what='calls' cc='IP02' showclosed='yes'></inesc-id>--><br />
<!-- PLEASE POST NEW POSITION ANNOUNCEMENTS THROUGH www.inesc-id.pt --><br />
<br />
== Ph.D. Program CMU-Portugal ==<br />
<br />
* [[Ph.D. Program Carnegie Mellon - PORTUGAL in the area of Language and Information Technologies|'''Ph.D. Program Carnegie Mellon-Portugal''' in the area of Language and Information Technologies]]<br />
<br />
== Post-Doc Fellowship in Project INSIDE ==<br />
<br />
This is a 3 year position, starting in April 2015.<br />
<br />
INSIDE is a joint research project involving Portuguese academic/industrial partners and Carnegie Mellon University. The project explores symbiotic child-robot interactions in joint cooperative activities, with application to the treatment of children with Autism Spectrum Disorder impairments.<br />
<br />
Your role would be related to speech processing in environmentally challenging conditions and with limited adaptation data available. Research will be focused on static and dynamic adaptation of models starting from our current research on deep learning and observation uncertainty techniques.<br />
<br />
We are looking for candidates with following experience:<br />
<br />
* Robust ASR/KWS<br />
* Neural networks<br />
* Python, bash scripting and C/C++<br />
<br />
Additional experience welcome<br />
<br />
* Computational Paralinguistics<br />
* Python-theano<br />
* Knowledge of the Portuguese language<br />
* Missing features / Observation uncertainty<br />
<br />
If interested, send a curriculum vitae (highlighting publications relevant to the position) and a motivation letter to ramon.astudillo@l2f.inesc-id.pt before March 15.<br />
<br />
<br />
<br />
== Opportunities for Final Year Students ==<br />
<br />
(currently closed)<br />
<br />
== Other Information ==<br />
<br />
* A [[Positions|complete list]] of calls posted on [http://www.inesc-id.pt/intranet/bolsas/list_fellowships.php INESC ID's website], with both open and closed calls.<br />
* [[Older Calls]]</div>Ramonhttps://www.hlt.inesc-id.pt/wiki/index.php?title=Opportunities_(all_announcements)&diff=7524Opportunities (all announcements)2015-02-23T20:12:50Z<p>Ramon: </p>
<hr />
<div>__NOTOC__<br />
== Thinking about Joining Us? ==<br />
INESC ID is one of the largest and most dynamic research institutes in Portugal in the areas of telecommunications, information systems and electronics with the status of associated laboratory given by the Portuguese government. The Laboratorio de Lengua Falada (L2F), or Spoken Language Systems Lab in English, provides an interdisciplinary and creative work environment active in multiple areas of speech and language research. This include <br />
<br />
* speech enhancement, recognition, synthesis and understanding, speech to speech translation<br />
<br />
* speaker, language and dialect identification, computational paralinguistics<br />
<br />
* music genre classification, audio finger-printing and retrieval, multi-document summarization of multimedia (audio + video + text)<br />
<br />
* machine translation, summarization, sentiment analysis<br />
<br />
* natural language generation, syntactic and semantic processing, named entity retrieval<br />
<br />
* multimodal dialogue systems, question answering, <br />
<br />
<br />
L2F has its own powerful IT infrastructure within INESC-ID including <br />
<br />
* Robust and fast shared filesystems for storage and processing of very large data-sets<br />
<br />
* A computer cluster with Hadoop and condor among other parallel computing utilities <br />
<br />
* Dedicated computing units with GPUs and intel-phi CPUs<br />
<br />
* VPN, ssh as well as hosting for CVS, SVN and Git repositories.<br />
<br />
This infrastructure also provides online services to various academic and private sector partners. We also share resources with our spin-off [http://www.voiceinteraction.pt/ VoiceInteraction] and cooperate with well established industrial partners such as [http://www.microsoft.com/pt-pt/mldc/ Microsoft] as well as startups such as [https://unbabel.com/ Unbabel]. <br />
<br />
<br />
Have a look at our recent [https://www.l2f.inesc-id.pt/wiki/index.php/Projects projects] and [https://www.l2f.inesc-id.pt/wiki/index.php/Demos demos]. <br />
<br />
Want to joint us?. Have a look at the opportunities below. <br />
<br />
<!-- Ramon: I commented this because it did not update --><br />
<!-- <inesc-id what='calls' cc='IP02'></inesc-id> --><br />
<!-- <inesc-id what='calls' cc='IP02' showclosed='yes'></inesc-id>--><br />
<!-- PLEASE POST NEW POSITION ANNOUNCEMENTS THROUGH www.inesc-id.pt --><br />
<br />
== Ph.D. Program CMU-Portugal ==<br />
<br />
* [[Ph.D. Program Carnegie Mellon - PORTUGAL in the area of Language and Information Technologies|'''Ph.D. Program Carnegie Mellon-Portugal''' in the area of Language and Information Technologies]]<br />
<br />
== Post-Doc Fellowship in Project INSIDE ==<br />
<br />
This is a 3 year position, starting in April 2015.<br />
<br />
INSIDE is a joint research project involving Portuguese academic/industrial partners and Carnegie Mellon University. The project explores symbiotic child-robot interactions in joint cooperative activities, with application to the treatment of children with Autism Spectrum Disorder impairments.<br />
<br />
Your role would be related to speech processing in environmentally challenging conditions and with limited adaptation data available. Research will be focused on static and dynamic adaptation of models starting from our current research on deep learning and observation uncertainty techniques.<br />
<br />
We are looking for candidates with following experience:<br />
<br />
* Robust ASR/KWS<br />
* Neural networks<br />
* Python, bash scripting and C/C++<br />
<br />
Additional experience welcome<br />
<br />
* Computational Paralinguistics<br />
* Python-theano<br />
* Knowledge of the Portuguese language<br />
* Missing features / Observation uncertainty<br />
<br />
If interested, send a curriculum vitae (highlighting publications relevant to the position) and a motivation letter to ramon.astudillo@l2f.inesc-id.pt before March 15.<br />
<br />
<br />
<br />
== Opportunities for Final Year Students ==<br />
<br />
(currently closed)<br />
<br />
== Other Information ==<br />
<br />
* A [[Positions|complete list]] of calls posted on [http://www.inesc-id.pt/intranet/bolsas/list_fellowships.php INESC ID's website], with both open and closed calls.<br />
* [[Older Calls]]</div>Ramonhttps://www.hlt.inesc-id.pt/wiki/index.php?title=Opportunities_(all_announcements)&diff=7523Opportunities (all announcements)2015-02-23T20:12:36Z<p>Ramon: </p>
<hr />
<div>__NOTOC__<br />
== Thinking about Joining Us? ==<br />
INESC ID is one of the largest and most dynamic research institutes in Portugal in the areas of telecommunications, information systems and electronics with the status of associated laboratory given by the Portuguese government. The Laboratorio de Lengua Falada (L2F), or Spoken Language Systems Lab in English, provides an interdisciplinary and creative work environment active in multiple areas of speech and language research. This include <br />
<br />
* speech enhancement, recognition, synthesis and understanding, speech to speech translation<br />
<br />
* speaker, language and dialect identification, computational paralinguistics<br />
<br />
* music genre classification, audio finger-printing and retrieval, multi-document summarization of multimedia (audio + video + text)<br />
<br />
<br />
* machine translation, summarization, sentiment analysis<br />
<br />
* natural language generation, syntactic and semantic processing, named entity retrieval<br />
<br />
* multimodal dialogue systems, question answering, <br />
<br />
<br />
L2F has its own powerful IT infrastructure within INESC-ID including <br />
<br />
* Robust and fast shared filesystems for storage and processing of very large data-sets<br />
<br />
* A computer cluster with Hadoop and condor among other parallel computing utilities <br />
<br />
* Dedicated computing units with GPUs and intel-phi CPUs<br />
<br />
* VPN, ssh as well as hosting for CVS, SVN and Git repositories.<br />
<br />
This infrastructure also provides online services to various academic and private sector partners. We also share resources with our spin-off [http://www.voiceinteraction.pt/ VoiceInteraction] and cooperate with well established industrial partners such as [http://www.microsoft.com/pt-pt/mldc/ Microsoft] as well as startups such as [https://unbabel.com/ Unbabel]. <br />
<br />
<br />
Have a look at our recent [https://www.l2f.inesc-id.pt/wiki/index.php/Projects projects] and [https://www.l2f.inesc-id.pt/wiki/index.php/Demos demos]. <br />
<br />
Want to joint us?. Have a look at the opportunities below. <br />
<br />
<!-- Ramon: I commented this because it did not update --><br />
<!-- <inesc-id what='calls' cc='IP02'></inesc-id> --><br />
<!-- <inesc-id what='calls' cc='IP02' showclosed='yes'></inesc-id>--><br />
<!-- PLEASE POST NEW POSITION ANNOUNCEMENTS THROUGH www.inesc-id.pt --><br />
<br />
== Ph.D. Program CMU-Portugal ==<br />
<br />
* [[Ph.D. Program Carnegie Mellon - PORTUGAL in the area of Language and Information Technologies|'''Ph.D. Program Carnegie Mellon-Portugal''' in the area of Language and Information Technologies]]<br />
<br />
== Post-Doc Fellowship in Project INSIDE ==<br />
<br />
This is a 3 year position, starting in April 2015.<br />
<br />
INSIDE is a joint research project involving Portuguese academic/industrial partners and Carnegie Mellon University. The project explores symbiotic child-robot interactions in joint cooperative activities, with application to the treatment of children with Autism Spectrum Disorder impairments.<br />
<br />
Your role would be related to speech processing in environmentally challenging conditions and with limited adaptation data available. Research will be focused on static and dynamic adaptation of models starting from our current research on deep learning and observation uncertainty techniques.<br />
<br />
We are looking for candidates with following experience:<br />
<br />
* Robust ASR/KWS<br />
* Neural networks<br />
* Python, bash scripting and C/C++<br />
<br />
Additional experience welcome<br />
<br />
* Computational Paralinguistics<br />
* Python-theano<br />
* Knowledge of the Portuguese language<br />
* Missing features / Observation uncertainty<br />
<br />
If interested, send a curriculum vitae (highlighting publications relevant to the position) and a motivation letter to ramon.astudillo@l2f.inesc-id.pt before March 15.<br />
<br />
<br />
<br />
== Opportunities for Final Year Students ==<br />
<br />
(currently closed)<br />
<br />
== Other Information ==<br />
<br />
* A [[Positions|complete list]] of calls posted on [http://www.inesc-id.pt/intranet/bolsas/list_fellowships.php INESC ID's website], with both open and closed calls.<br />
* [[Older Calls]]</div>Ramonhttps://www.hlt.inesc-id.pt/wiki/index.php?title=Opportunities_(all_announcements)&diff=7522Opportunities (all announcements)2015-02-23T20:09:03Z<p>Ramon: </p>
<hr />
<div>__NOTOC__<br />
== Join Us ==<br />
INESC ID is one of the largest and most dynamic research institutes in Portugal in the areas of telecommunications, information systems and electronics with the status of associated laboratory given by the Portuguese government. The Laboratorio de Lengua Falada (L2F), or Spoken Language Systems Lab in English, provides an interdisciplinary and creative work environment active in multiple areas of speech and language research. This include <br />
<br />
* speech enhancement, recognition, synthesis and understanding, speech to speech translation<br />
<br />
* speaker, language and dialect identification, computational paralinguistics<br />
<br />
* music genre classification, audio finger-printing and retrieval, multi-document summarization of multimedia (audio + video + text)<br />
<br />
<br />
* machine translation, summarization, sentiment analysis<br />
<br />
* natural language generation, syntactic and semantic processing, named entity retrieval<br />
<br />
* multimodal dialogue systems, question answering, <br />
<br />
<br />
L2F has its own powerful IT infrastructure within INESC-ID including <br />
<br />
* Robust and fast shared filesystems for storage and processing of very large data-sets<br />
<br />
* A computer cluster with Hadoop and condor among other parallel computing utilities <br />
<br />
* Dedicated computing units with GPUs and intel-phi CPUs<br />
<br />
* VPN, ssh as well as hosting for CVS, SVN and Git repositories.<br />
<br />
This infrastructure also provides online services to various academic and private sector partners. We also share resources with our spin-off [http://www.voiceinteraction.pt/ VoiceInteraction] and cooperate with well established industrial partners such as [http://www.microsoft.com/pt-pt/mldc/ Microsoft] as well as startups such as [https://unbabel.com/ Unbabel]. <br />
<br />
<br />
Have a look at our recent [https://www.l2f.inesc-id.pt/wiki/index.php/Projects projects] and [https://www.l2f.inesc-id.pt/wiki/index.php/Demos demos]. <br />
<br />
Want to joint us?. Have a look at the opportunities below. <br />
<br />
<!-- Ramon: I commented this because it did not update --><br />
<!-- <inesc-id what='calls' cc='IP02'></inesc-id> --><br />
<!-- <inesc-id what='calls' cc='IP02' showclosed='yes'></inesc-id>--><br />
<!-- PLEASE POST NEW POSITION ANNOUNCEMENTS THROUGH www.inesc-id.pt --><br />
<br />
== Ph.D. Program CMU-Portugal ==<br />
<br />
* [[Ph.D. Program Carnegie Mellon - PORTUGAL in the area of Language and Information Technologies|'''Ph.D. Program Carnegie Mellon-Portugal''' in the area of Language and Information Technologies]]<br />
<br />
== Post-Doc Fellowship in Project INSIDE ==<br />
<br />
This is a 3 year position, starting in April 2015.<br />
<br />
INSIDE is a joint research project involving Portuguese academic/industrial partners and Carnegie Mellon University. The project explores symbiotic child-robot interactions in joint cooperative activities, with application to the treatment of children with Autism Spectrum Disorder impairments.<br />
<br />
Your role would be related to speech processing in environmentally challenging conditions and with limited adaptation data available. Research will be focused on static and dynamic adaptation of models starting from our current research on deep learning and observation uncertainty techniques.<br />
<br />
We are looking for candidates with following experience:<br />
<br />
* Robust ASR/KWS<br />
* Neural networks<br />
* Python, bash scripting and C/C++<br />
<br />
Additional experience welcome<br />
<br />
* Computational Paralinguistics<br />
* Python-theano<br />
* Knowledge of the Portuguese language<br />
* Missing features / Observation uncertainty<br />
<br />
If interested, send a curriculum vitae (highlighting publications relevant to the position) and a motivation letter to ramon.astudillo@l2f.inesc-id.pt before March 15.<br />
<br />
<br />
<br />
== Opportunities for Final Year Students ==<br />
<br />
(currently closed)<br />
<br />
== Other Information ==<br />
<br />
* A [[Positions|complete list]] of calls posted on [http://www.inesc-id.pt/intranet/bolsas/list_fellowships.php INESC ID's website], with both open and closed calls.<br />
* [[Older Calls]]</div>Ramonhttps://www.hlt.inesc-id.pt/wiki/index.php?title=Opportunities_(all_announcements)&diff=7521Opportunities (all announcements)2015-02-23T19:39:05Z<p>Ramon: </p>
<hr />
<div>__NOTOC__<br />
== Join Us ==<br />
[http://www.inesc-id.pt/ INESC ID] is one of the largest and most dynamic research institutes in Portugal in the areas of telecommunications, information systems and electronics with the status of associated laboratory given by the Portuguese government. The Laboratorio de Lengua Falada (L2F), or Spoken Language Systems Lab in English, provides an interdisciplinary and creative work environment active in multiple areas of speech and language research. This include <br />
<br />
* speech enhancement, recognition, synthesis and understanding, speech to speech translation<br />
<br />
* speaker, language and dialect identification, computational paralinguistics<br />
<br />
* music genre classification, audio finger-printing and retrieval, multi-document summarization of multimedia (audio + video + text)<br />
<br />
<br />
* machine translation, summarization, sentiment analysis<br />
<br />
* natural language generation, syntactic and semantic processing, named entity retrieval<br />
<br />
* multimodal dialogue systems, question answering, <br />
<br />
<br />
L2F has its own powerful IT infraesctructure within INESC-ID including <br />
<br />
* Robust and fast shared filesystems for storage and processing of very large data-sets<br />
<br />
* A computer cluster with Hadoop and condor among other parallel computing utilities <br />
<br />
* Dedicated computing units with GPUs and intel-phi CPUs<br />
<br />
* VPN, ssh as well as hosting for CVS, SVN and Git repositories.<br />
<br />
This infrastructure also provides online services to various academic and private sector partners.<br />
<br />
In addition to this, L2F <br />
<br />
<br />
<br />
Have a look at our recent [https://www.l2f.inesc-id.pt/wiki/index.php/Projects projects] and [https://www.l2f.inesc-id.pt/wiki/index.php/Demos demos]. <br />
<br />
Want to joint us?. Have a look at the opportunities below. <br />
<br />
<!-- Ramon: I commented this because it did not update --><br />
<!-- <inesc-id what='calls' cc='IP02'></inesc-id> --><br />
<!-- <inesc-id what='calls' cc='IP02' showclosed='yes'></inesc-id>--><br />
<!-- PLEASE POST NEW POSITION ANNOUNCEMENTS THROUGH www.inesc-id.pt --><br />
<br />
== Ph.D. Program CMU-Portugal ==<br />
<br />
* [[Ph.D. Program Carnegie Mellon - PORTUGAL in the area of Language and Information Technologies|'''Ph.D. Program Carnegie Mellon-Portugal''' in the area of Language and Information Technologies]]<br />
<br />
== Post-Doc Fellowship in Project INSIDE ==<br />
<br />
This is a 3 year position, starting in April 2015.<br />
<br />
INSIDE is a joint research project involving Portuguese academic/industrial partners and Carnegie Mellon University. The project explores symbiotic child-robot interactions in joint cooperative activities, with application to the treatment of children with Autism Spectrum Disorder impairments.<br />
<br />
Your role would be related to speech processing in environmentally challenging conditions and with limited adaptation data available. Research will be focused on static and dynamic adaptation of models starting from our current research on deep learning and observation uncertainty techniques.<br />
<br />
We are looking for candidates with following experience:<br />
<br />
* Robust ASR/KWS<br />
* Neural networks<br />
* Python, bash scripting and C/C++<br />
<br />
Additional experience welcome<br />
<br />
* Computational Paralinguistics<br />
* Python-theano<br />
* Knowledge of the Portuguese language<br />
* Missing features / Observation uncertainty<br />
<br />
If interested, send a curriculum vitae (highlighting publications relevant to the position) and a motivation letter to ramon.astudillo@l2f.inesc-id.pt before March 15.<br />
<br />
<br />
<br />
== Opportunities for Final Year Students ==<br />
<br />
(currently closed)<br />
<br />
== Other Information ==<br />
<br />
* A [[Positions|complete list]] of calls posted on [http://www.inesc-id.pt/intranet/bolsas/list_fellowships.php INESC ID's website], with both open and closed calls.<br />
* [[Older Calls]]</div>Ramonhttps://www.hlt.inesc-id.pt/wiki/index.php?title=Opportunities_(all_announcements)&diff=7520Opportunities (all announcements)2015-02-23T18:55:24Z<p>Ramon: </p>
<hr />
<div>__NOTOC__<br />
== Join Us ==<br />
[http://www.inesc-id.pt/ INESC ID] is one of the largest and most dynamic research institutes in Portugal in the areas of telecommunications, information systems and electronics with the status of associated laboratory given by the Portuguese government. The Laboratorio de Lengua Falada (L2F), or Spoken Language Systems Lab in English, provides an interdisciplinary and creative work environment active in multiple areas of speech and language research. This include <br />
<br />
* speech enhancement, recognition, synthesis and understanding, speech to speech translation<br />
<br />
* speaker, language and dialect identification, computational paralinguistics<br />
<br />
* music genre classification, audio finger-printing and retrieval <br />
<br />
* machine translation, summarization, sentiment analysis<br />
<br />
* natural language generation, syntactic and semantic processing, named entity retrieval<br />
<br />
* multimodal dialogue systems, question answering, <br />
<br />
<br />
L2F has its own powerful IT infraesctructure within INESC-ID including <br />
<br />
* Robust and fast shared filesystems for storage and processing of very large data-sets<br />
<br />
* A computer cluster with Hadoop and condor among other parallel computing utilities <br />
<br />
* Dedicated computing units with GPUs and intel-phi CPUs<br />
<br />
* VPN, ssh as well as hosting for CVS, SVN and Git repositories.<br />
<br />
This infrastructure also provides online services to various academic and private sector partners.<br />
<br />
In addition to this L2F, is proud to<br />
<br />
<br />
<br />
Have a look at our recent [https://www.l2f.inesc-id.pt/wiki/index.php/Projects projects] and [https://www.l2f.inesc-id.pt/wiki/index.php/Demos demos]. <br />
<br />
Want to joint us?. Have a look at the opportunities below. <br />
<br />
<!-- Ramon: I commented this because it did not update --><br />
<!-- <inesc-id what='calls' cc='IP02'></inesc-id> --><br />
<!-- <inesc-id what='calls' cc='IP02' showclosed='yes'></inesc-id>--><br />
<!-- PLEASE POST NEW POSITION ANNOUNCEMENTS THROUGH www.inesc-id.pt --><br />
<br />
== Ph.D. Program CMU-Portugal ==<br />
<br />
* [[Ph.D. Program Carnegie Mellon - PORTUGAL in the area of Language and Information Technologies|'''Ph.D. Program Carnegie Mellon-Portugal''' in the area of Language and Information Technologies]]<br />
<br />
== Post-Doc Fellowship in Project INSIDE ==<br />
<br />
This is a 3 year position, starting in April 2015.<br />
<br />
INSIDE is a joint research project involving Portuguese academic/industrial partners and Carnegie Mellon University. The project explores symbiotic child-robot interactions in joint cooperative activities, with application to the treatment of children with Autism Spectrum Disorder impairments.<br />
<br />
Your role would be related to speech processing in environmentally challenging conditions and with limited adaptation data available. Research will be focused on static and dynamic adaptation of models starting from our current research on deep learning and observation uncertainty techniques.<br />
<br />
We are looking for candidates with following experience:<br />
<br />
* Robust ASR/KWS<br />
* Neural networks<br />
* Python, bash scripting and C/C++<br />
<br />
Additional experience welcome<br />
<br />
* Computational Paralinguistics<br />
* Python-theano<br />
* Knowledge of the Portuguese language<br />
* Missing features / Observation uncertainty<br />
<br />
If interested, send a curriculum vitae (highlighting publications relevant to the position) and a motivation letter to ramon.astudillo@l2f.inesc-id.pt before March 15.<br />
<br />
<br />
<br />
== Opportunities for Final Year Students ==<br />
<br />
(currently closed)<br />
<br />
== Other Information ==<br />
<br />
* A [[Positions|complete list]] of calls posted on [http://www.inesc-id.pt/intranet/bolsas/list_fellowships.php INESC ID's website], with both open and closed calls.<br />
* [[Older Calls]]</div>Ramonhttps://www.hlt.inesc-id.pt/wiki/index.php?title=Opportunities_(all_announcements)&diff=7519Opportunities (all announcements)2015-02-23T18:42:49Z<p>Ramon: </p>
<hr />
<div>__NOTOC__<br />
== Join Us ==<br />
[http://www.inesc-id.pt/ INESC ID] is one of the largest and most dynamic research institutes in Portugal in the areas of telecommunications, information systems and electronics with the status of associated laboratory given by the Portuguese government. The Laboratorio de Lengua Falada (L2F), or Spoken Language Systems Lab in English, provides an interdisciplinary and creative work environment active in multiple areas of speech and language research. This include <br />
<br />
* speech enhancement, recognition, synthesis and understanding, speech to speech translation<br />
<br />
* speaker, language and dialect identification, computational paralinguistics<br />
<br />
* music genre classification, audio finger-printing and retrieval <br />
<br />
* machine translation, summarization, sentiment analysis<br />
<br />
* natural language generation, syntactic and semantic processing, named entity retrieval<br />
<br />
* multimodal dialogue systems, question answering, <br />
<br />
<br />
L2F has its own powerful IT infraesctructure within INESC-ID including <br />
<br />
* Robust and fast shared filesystems for storage and processing of very large data-sets<br />
<br />
* A computer cluster with Hadoop and condor among other parallel computing utilities <br />
<br />
* Dedicated computing units with GPUs and intel-phi CPUs<br />
<br />
* VPN, ssh as well as hosting for CVS, SVN and Git repositories.<br />
<br />
This infrastructure also provides online services to various academic and private sector partners.<br />
<br />
<br />
<br />
Have a look at our recent [https://www.l2f.inesc-id.pt/wiki/index.php/Projects projects] and [https://www.l2f.inesc-id.pt/wiki/index.php/Demos demos]. <br />
<br />
Want to joint us?. Have a look at the opportunities below. <br />
<br />
<!-- Ramon: I commented this because it did not update --><br />
<!-- <inesc-id what='calls' cc='IP02'></inesc-id> --><br />
<!-- <inesc-id what='calls' cc='IP02' showclosed='yes'></inesc-id>--><br />
<!-- PLEASE POST NEW POSITION ANNOUNCEMENTS THROUGH www.inesc-id.pt --><br />
<br />
== Ph.D. Program CMU-Portugal ==<br />
<br />
* [[Ph.D. Program Carnegie Mellon - PORTUGAL in the area of Language and Information Technologies|'''Ph.D. Program Carnegie Mellon-Portugal''' in the area of Language and Information Technologies]]<br />
<br />
== Post-Doc Fellowship in Project INSIDE ==<br />
<br />
This is a 3 year position, starting in April 2015.<br />
<br />
INSIDE is a joint research project involving Portuguese academic/industrial partners and Carnegie Mellon University. The project explores symbiotic child-robot interactions in joint cooperative activities, with application to the treatment of children with Autism Spectrum Disorder impairments.<br />
<br />
Your role would be related to speech processing in environmentally challenging conditions and with limited adaptation data available. Research will be focused on static and dynamic adaptation of models starting from our current research on deep learning and observation uncertainty techniques.<br />
<br />
We are looking for candidates with following experience:<br />
<br />
* Robust ASR/KWS<br />
* Neural networks<br />
* Python, bash scripting and C/C++<br />
<br />
Additional experience welcome<br />
<br />
* Computational Paralinguistics<br />
* Python-theano<br />
* Knowledge of the Portuguese language<br />
* Missing features / Observation uncertainty<br />
<br />
If interested, send a curriculum vitae (highlighting publications relevant to the position) and a motivation letter to ramon.astudillo@l2f.inesc-id.pt before March 15.<br />
<br />
<br />
<br />
== Opportunities for Final Year Students ==<br />
<br />
(currently closed)<br />
<br />
== Other Information ==<br />
<br />
* A [[Positions|complete list]] of calls posted on [http://www.inesc-id.pt/intranet/bolsas/list_fellowships.php INESC ID's website], with both open and closed calls.<br />
* [[Older Calls]]</div>Ramonhttps://www.hlt.inesc-id.pt/wiki/index.php?title=Opportunities_(all_announcements)&diff=7518Opportunities (all announcements)2015-02-23T18:39:59Z<p>Ramon: </p>
<hr />
<div>__NOTOC__<br />
== Join Us ==<br />
[http://www.inesc-id.pt/ INESC ID] is one of the largest and most dynamic research institutes in Portugal in the areas of telecommunications, information systems and electronics with the status of associated laboratory given by the Portuguese government. The Laboratorio de Lengua Falada (L2F), or Spoken Language Systems Lab in English, provides an interdisciplinary and creative work environment active in multiple areas of speech and language research. This include <br />
<br />
* speech enhancement, recognition, synthesis and understanding, speech to speech translation<br />
<br />
* speaker, language and dialect identification, computational paralinguistics<br />
<br />
* music genre classification, audio finger-printing and retrieval <br />
<br />
* machine translation, summarization<br />
<br />
* natural language generation, syntactic and semantic processing, named entity retrieval<br />
<br />
* multimodal dialogue systems, question answering, <br />
<br />
<br />
L2F has its own powerful IT infraesctructure within INESC-ID including <br />
<br />
* Robust and fast shared filesystems for storage and processing of very large data-sets<br />
<br />
* A computer cluster with Hadoop and condor among other parallel computing utilities <br />
<br />
* Dedicated computing units with GPUs and intel-phi CPUs<br />
<br />
* VPN, ssh as well as hosting for CVS, SVN and Git repositories.<br />
<br />
This infrastructure also provides online services to various academic and private sector partners.<br />
<br />
<br />
<br />
Have a look at our recent [https://www.l2f.inesc-id.pt/wiki/index.php/Projects projects] and [https://www.l2f.inesc-id.pt/wiki/index.php/Demos demos]. <br />
<br />
Want to joint us?. Have a look at the opportunities below. <br />
<br />
<!-- Ramon: I commented this because it did not update --><br />
<!-- <inesc-id what='calls' cc='IP02'></inesc-id> --><br />
<!-- <inesc-id what='calls' cc='IP02' showclosed='yes'></inesc-id>--><br />
<!-- PLEASE POST NEW POSITION ANNOUNCEMENTS THROUGH www.inesc-id.pt --><br />
<br />
== Ph.D. Program CMU-Portugal ==<br />
<br />
* [[Ph.D. Program Carnegie Mellon - PORTUGAL in the area of Language and Information Technologies|'''Ph.D. Program Carnegie Mellon-Portugal''' in the area of Language and Information Technologies]]<br />
<br />
== Post-Doc Fellowship in Project INSIDE ==<br />
<br />
This is a 3 year position, starting in April 2015.<br />
<br />
INSIDE is a joint research project involving Portuguese academic/industrial partners and Carnegie Mellon University. The project explores symbiotic child-robot interactions in joint cooperative activities, with application to the treatment of children with Autism Spectrum Disorder impairments.<br />
<br />
Your role would be related to speech processing in environmentally challenging conditions and with limited adaptation data available. Research will be focused on static and dynamic adaptation of models starting from our current research on deep learning and observation uncertainty techniques.<br />
<br />
We are looking for candidates with following experience:<br />
<br />
* Robust ASR/KWS<br />
* Neural networks<br />
* Python, bash scripting and C/C++<br />
<br />
Additional experience welcome<br />
<br />
* Computational Paralinguistics<br />
* Python-theano<br />
* Knowledge of the Portuguese language<br />
* Missing features / Observation uncertainty<br />
<br />
If interested, send a curriculum vitae (highlighting publications relevant to the position) and a motivation letter to ramon.astudillo@l2f.inesc-id.pt before March 15.<br />
<br />
<br />
<br />
== Opportunities for Final Year Students ==<br />
<br />
(currently closed)<br />
<br />
== Other Information ==<br />
<br />
* A [[Positions|complete list]] of calls posted on [http://www.inesc-id.pt/intranet/bolsas/list_fellowships.php INESC ID's website], with both open and closed calls.<br />
* [[Older Calls]]</div>Ramonhttps://www.hlt.inesc-id.pt/wiki/index.php?title=Opportunities_(all_announcements)&diff=7517Opportunities (all announcements)2015-02-23T16:45:50Z<p>Ramon: </p>
<hr />
<div>__NOTOC__<br />
== Join Us ==<br />
[http://www.inesc-id.pt/ INESC ID] is one of the largest and most dynamic research institutes in Portugal in the areas of telecommunications, information systems and electronics with the status of associated laboratory given by the Portuguese government. The Laboratorio de Lengua Falada (L2F), or Spoken Language Systems Lab in English, provides an interdisciplinary and creative work environment active in multiple areas of speech and language research. This includes speech recognition, synthesis, and understanding, audio and music processing, multimodal dialogue systems, language and dialect identification, speech-to-speech machine translation, natural language generation, syntactic and semantic processing, summarization, named entity retrieval, question answering and more. We also have a powerful IT architecture including robust and fast shared filesystems, a computer cluster (with Hadoop, condor) specialized computing units (GPUs, intel-phi) as well as VPN access. This infrastructure also provides services to academic and private sector partners.<br />
<br />
Have a look at our recent [https://www.l2f.inesc-id.pt/wiki/index.php/Projects projects] and [https://www.l2f.inesc-id.pt/wiki/index.php/Demos demos]. <br />
<br />
Want to joint us?. Have a look at the opportunities below. <br />
<br />
<!-- Ramon: I commented this because it did not update --><br />
<!-- <inesc-id what='calls' cc='IP02'></inesc-id> --><br />
<!-- <inesc-id what='calls' cc='IP02' showclosed='yes'></inesc-id>--><br />
<!-- PLEASE POST NEW POSITION ANNOUNCEMENTS THROUGH www.inesc-id.pt --><br />
<br />
== Ph.D. Program CMU-Portugal ==<br />
<br />
* [[Ph.D. Program Carnegie Mellon - PORTUGAL in the area of Language and Information Technologies|'''Ph.D. Program Carnegie Mellon-Portugal''' in the area of Language and Information Technologies]]<br />
<br />
== Post-Doc Fellowship in Project INSIDE ==<br />
<br />
This is a 3 year position, starting in April 2015.<br />
<br />
INSIDE is a joint research project involving Portuguese academic/industrial partners and Carnegie Mellon University. The project explores symbiotic child-robot interactions in joint cooperative activities, with application to the treatment of children with Autism Spectrum Disorder impairments.<br />
<br />
Your role would be related to speech processing in environmentally challenging conditions and with limited adaptation data available. Research will be focused on static and dynamic adaptation of models starting from our current research on deep learning and observation uncertainty techniques.<br />
<br />
We are looking for candidates with following experience:<br />
<br />
* Robust ASR/KWS<br />
* Neural networks<br />
* Python, bash scripting and C/C++<br />
<br />
Additional experience welcome<br />
<br />
* Computational Paralinguistics<br />
* Python-theano<br />
* Knowledge of the Portuguese language<br />
* Missing features / Observation uncertainty<br />
<br />
If interested, send a curriculum vitae (highlighting publications relevant to the position) and a motivation letter to ramon.astudillo@l2f.inesc-id.pt before March 15.<br />
<br />
<br />
<br />
== Opportunities for Final Year Students ==<br />
<br />
(currently closed)<br />
<br />
== Other Information ==<br />
<br />
* A [[Positions|complete list]] of calls posted on [http://www.inesc-id.pt/intranet/bolsas/list_fellowships.php INESC ID's website], with both open and closed calls.<br />
* [[Older Calls]]</div>Ramonhttps://www.hlt.inesc-id.pt/wiki/index.php?title=Opportunities_(all_announcements)&diff=7516Opportunities (all announcements)2015-02-23T16:45:22Z<p>Ramon: </p>
<hr />
<div>__NOTOC__<br />
== Join Us ==<br />
[http://www.inesc-id.pt/ INESC ID] is one of the largest and most dynamic research institutes in Portugal in the areas of telecommunications, information systems and electronics with the status of associated laboratory given by the [http://www.portugal.gov.pt/ Portuguese government]. <br />
<br />
The Laboratorio de Lengua Falada (L2F), or Spoken Language Systems Lab in English, provides an interdisciplinary and creative work environment active in multiple areas of speech and language research. This includes speech recognition, synthesis, and understanding, audio and music processing, multimodal dialogue systems, language and dialect identification, speech-to-speech machine translation, natural language generation, syntactic and semantic processing, summarization, named entity retrieval, question answering and more. We also have a powerful IT architecture including robust and fast shared filesystems, a computer cluster (with Hadoop, condor) specialized computing units (GPUs, intel-phi) as well as VPN access. This infrastructure also provides services to academic and private sector partners.<br />
<br />
Have a look at our recent [https://www.l2f.inesc-id.pt/wiki/index.php/Projects projects] and [https://www.l2f.inesc-id.pt/wiki/index.php/Demos demos]. <br />
<br />
Want to joint us?. Have a look at the opportunities below. <br />
<br />
<!-- Ramon: I commented this because it did not update --><br />
<!-- <inesc-id what='calls' cc='IP02'></inesc-id> --><br />
<!-- <inesc-id what='calls' cc='IP02' showclosed='yes'></inesc-id>--><br />
<!-- PLEASE POST NEW POSITION ANNOUNCEMENTS THROUGH www.inesc-id.pt --><br />
<br />
== Ph.D. Program CMU-Portugal ==<br />
<br />
* [[Ph.D. Program Carnegie Mellon - PORTUGAL in the area of Language and Information Technologies|'''Ph.D. Program Carnegie Mellon-Portugal''' in the area of Language and Information Technologies]]<br />
<br />
== Post-Doc Fellowship in Project INSIDE ==<br />
<br />
This is a 3 year position, starting in April 2015.<br />
<br />
INSIDE is a joint research project involving Portuguese academic/industrial partners and Carnegie Mellon University. The project explores symbiotic child-robot interactions in joint cooperative activities, with application to the treatment of children with Autism Spectrum Disorder impairments.<br />
<br />
Your role would be related to speech processing in environmentally challenging conditions and with limited adaptation data available. Research will be focused on static and dynamic adaptation of models starting from our current research on deep learning and observation uncertainty techniques.<br />
<br />
We are looking for candidates with following experience:<br />
<br />
* Robust ASR/KWS<br />
* Neural networks<br />
* Python, bash scripting and C/C++<br />
<br />
Additional experience welcome<br />
<br />
* Computational Paralinguistics<br />
* Python-theano<br />
* Knowledge of the Portuguese language<br />
* Missing features / Observation uncertainty<br />
<br />
If interested, send a curriculum vitae (highlighting publications relevant to the position) and a motivation letter to ramon.astudillo@l2f.inesc-id.pt before March 15.<br />
<br />
<br />
<br />
== Opportunities for Final Year Students ==<br />
<br />
(currently closed)<br />
<br />
== Other Information ==<br />
<br />
* A [[Positions|complete list]] of calls posted on [http://www.inesc-id.pt/intranet/bolsas/list_fellowships.php INESC ID's website], with both open and closed calls.<br />
* [[Older Calls]]</div>Ramonhttps://www.hlt.inesc-id.pt/wiki/index.php?title=Opportunities_(all_announcements)&diff=7515Opportunities (all announcements)2015-02-23T16:15:04Z<p>Ramon: </p>
<hr />
<div>__NOTOC__<br />
[http://www.inesc-id.pt/ INESC ID] is one of the largest and most dynamic research institutes in Portugal in the areas of telecommunications, information systems and electronics with the status of associated laboratory given by the [http://www.portugal.gov.pt/ Portuguese government]. <br />
<br />
The Laboratorio de Lengua Falada (L2F), or Spoken Language Systems Lab in English, provides an interdisciplinary and creative work environment active in multiple areas of speech and language research. This includes speech recognition, synthesis, and understanding, audio and music processing, multimodal dialogue systems, language and dialect identification, speech-to-speech machine translation, natural language generation, syntactic and semantic processing, summarization, named entity retrieval, question answering and more. We also have a powerful IT architecture including robust and fast shared filesystems, a computer cluster (with Hadoop, condor) specialized computing units (GPUs, intel-phi) as well as VPN access. This infrastructure also provides services to academic and private sector partners.<br />
<br />
Have a look at our recent [https://www.l2f.inesc-id.pt/wiki/index.php/Projects projects] and [https://www.l2f.inesc-id.pt/wiki/index.php/Demos demos]. <br />
<br />
Want to joint us?. Have a look at the opportunities below. <br />
<br />
<!-- Ramon: I commented this because it did not update --><br />
<!-- <inesc-id what='calls' cc='IP02'></inesc-id> --><br />
<!-- <inesc-id what='calls' cc='IP02' showclosed='yes'></inesc-id>--><br />
<!-- PLEASE POST NEW POSITION ANNOUNCEMENTS THROUGH www.inesc-id.pt --><br />
<br />
== Ph.D. Program CMU-Portugal ==<br />
<br />
* [[Ph.D. Program Carnegie Mellon - PORTUGAL in the area of Language and Information Technologies|'''Ph.D. Program Carnegie Mellon-Portugal''' in the area of Language and Information Technologies]]<br />
<br />
== Post-Doc Fellowship in Project INSIDE ==<br />
<br />
This is a 3 year position, starting in April 2015.<br />
<br />
INSIDE is a joint research project involving Portuguese academic/industrial partners and Carnegie Mellon University. The project explores symbiotic child-robot interactions in joint cooperative activities, with application to the treatment of children with Autism Spectrum Disorder impairments.<br />
<br />
Your role would be related to speech processing in environmentally challenging conditions and with limited adaptation data available. Research will be focused on static and dynamic adaptation of models starting from our current research on deep learning and observation uncertainty techniques.<br />
<br />
We are looking for candidates with following experience:<br />
<br />
* Robust ASR/KWS<br />
* Neural networks<br />
* Python, bash scripting and C/C++<br />
<br />
Additional experience welcome<br />
<br />
* Computational Paralinguistics<br />
* Python-theano<br />
* Knowledge of the Portuguese language<br />
* Missing features / Observation uncertainty<br />
<br />
If interested, send a curriculum vitae (highlighting publications relevant to the position) and a motivation letter to ramon.astudillo@l2f.inesc-id.pt before March 15.<br />
<br />
<br />
<br />
== Opportunities for Final Year Students ==<br />
<br />
(currently closed)<br />
<br />
== Other Information ==<br />
<br />
* A [[Positions|complete list]] of calls posted on [http://www.inesc-id.pt/intranet/bolsas/list_fellowships.php INESC ID's website], with both open and closed calls.<br />
* [[Older Calls]]</div>Ramonhttps://www.hlt.inesc-id.pt/wiki/index.php?title=Opportunities_(all_announcements)&diff=7514Opportunities (all announcements)2015-02-23T16:11:39Z<p>Ramon: </p>
<hr />
<div>__NOTOC__<br />
[http://www.inesc-id.pt/ INESC ID] is one of the largest and most dynamic research institutes in Portugal in the areas of telecommunications, information systems and electronics with the status of associated laboratory given by the [http://www.portugal.gov.pt/ Portuguese government]. <br />
<br />
The Laboratorio de Lengua Falada (L2F), or Spoken Language Systems Lab in English, provides an interdisciplinary and creative work environment active in multiple areas of speech and language research. This includes speech recognition, synthesis, and understanding, audio and music processing, multimodal dialogue systems, language and dialect identification, speech-to-speech machine translation, natural language generation, syntactic and semantic processing, summarization, named entity retrieval, question answering and more. We also have a powerful IT architecture including robust and fast shared filesystems, a computer cluster (with Hadoop, condor) specialized computing units (GPUs, intel-phi) as well as VPN access. <br />
<br />
<br />
<!-- Ramon: I commented this because it did not update --><br />
<!-- <inesc-id what='calls' cc='IP02'></inesc-id> --><br />
<!-- <inesc-id what='calls' cc='IP02' showclosed='yes'></inesc-id>--><br />
<!-- PLEASE POST NEW POSITION ANNOUNCEMENTS THROUGH www.inesc-id.pt --><br />
<br />
== Ph.D. Program CMU-Portugal ==<br />
<br />
* [[Ph.D. Program Carnegie Mellon - PORTUGAL in the area of Language and Information Technologies|'''Ph.D. Program Carnegie Mellon-Portugal''' in the area of Language and Information Technologies]]<br />
<br />
== Post-Doc Fellowship in Project INSIDE ==<br />
<br />
This is a 3 year position, starting in April 2015.<br />
<br />
INSIDE is a joint research project involving Portuguese academic/industrial partners and Carnegie Mellon University. The project explores symbiotic child-robot interactions in joint cooperative activities, with application to the treatment of children with Autism Spectrum Disorder impairments.<br />
<br />
Your role would be related to speech processing in environmentally challenging conditions and with limited adaptation data available. Research will be focused on static and dynamic adaptation of models starting from our current research on deep learning and observation uncertainty techniques.<br />
<br />
We are looking for candidates with following experience:<br />
<br />
* Robust ASR/KWS<br />
* Neural networks<br />
* Python, bash scripting and C/C++<br />
<br />
Additional experience welcome<br />
<br />
* Computational Paralinguistics<br />
* Python-theano<br />
* Knowledge of the Portuguese language<br />
* Missing features / Observation uncertainty<br />
<br />
If interested, send a curriculum vitae (highlighting publications relevant to the position) and a motivation letter to ramon.astudillo@l2f.inesc-id.pt before March 15.<br />
<br />
<br />
<br />
== Opportunities for Final Year Students ==<br />
<br />
(currently closed)<br />
<br />
== Other Information ==<br />
<br />
* A [[Positions|complete list]] of calls posted on [http://www.inesc-id.pt/intranet/bolsas/list_fellowships.php INESC ID's website], with both open and closed calls.<br />
* [[Older Calls]]</div>Ramonhttps://www.hlt.inesc-id.pt/wiki/index.php?title=Opportunities_(all_announcements)&diff=7513Opportunities (all announcements)2015-02-23T16:10:59Z<p>Ramon: </p>
<hr />
<div>__NOTOC__<br />
[http://www.inesc-id.pt/ INESC ID] is one of the largest and most dynamic research institutes in Portugal in the areas of telecommunications, information systems and electronics with the status of associated laboratory given by the [http://www.portugal.gov.pt/ Portuguese government]. <br />
<br />
The Laboratorio de Lengua Falada (L2F), or Spoken Language Systems Lab in English, provides an interdisciplinary and creative work environment active in multiple areas of speech and language research. This includes speech recognition, synthesis, and understanding, audio and music processing, multimodal dialogue systems, language and dialect identification, speech-to-speech machine translation, natural language generation, syntactic and semantic processing, summarization, named entity retrieval, question answering and more.<br />
<br />
We also have a powerful IT architecture including robust and fast shared filesystems, a computer cluster (with Hadoop, condor) specialized computing units (GPUs, intel-phi) as well as VPN access. We also provide various external services to either academic partners or private institutions.<br />
<br />
<br />
<!-- Ramon: I commented this because it did not update --><br />
<!-- <inesc-id what='calls' cc='IP02'></inesc-id> --><br />
<!-- <inesc-id what='calls' cc='IP02' showclosed='yes'></inesc-id>--><br />
<!-- PLEASE POST NEW POSITION ANNOUNCEMENTS THROUGH www.inesc-id.pt --><br />
<br />
== Ph.D. Program CMU-Portugal ==<br />
<br />
* [[Ph.D. Program Carnegie Mellon - PORTUGAL in the area of Language and Information Technologies|'''Ph.D. Program Carnegie Mellon-Portugal''' in the area of Language and Information Technologies]]<br />
<br />
== Post-Doc Fellowship in Project INSIDE ==<br />
<br />
This is a 3 year position, starting in April 2015.<br />
<br />
INSIDE is a joint research project involving Portuguese academic/industrial partners and Carnegie Mellon University. The project explores symbiotic child-robot interactions in joint cooperative activities, with application to the treatment of children with Autism Spectrum Disorder impairments.<br />
<br />
Your role would be related to speech processing in environmentally challenging conditions and with limited adaptation data available. Research will be focused on static and dynamic adaptation of models starting from our current research on deep learning and observation uncertainty techniques.<br />
<br />
We are looking for candidates with following experience:<br />
<br />
* Robust ASR/KWS<br />
* Neural networks<br />
* Python, bash scripting and C/C++<br />
<br />
Additional experience welcome<br />
<br />
* Computational Paralinguistics<br />
* Python-theano<br />
* Knowledge of the Portuguese language<br />
* Missing features / Observation uncertainty<br />
<br />
If interested, send a curriculum vitae (highlighting publications relevant to the position) and a motivation letter to ramon.astudillo@l2f.inesc-id.pt before March 15.<br />
<br />
<br />
<br />
== Opportunities for Final Year Students ==<br />
<br />
(currently closed)<br />
<br />
== Other Information ==<br />
<br />
* A [[Positions|complete list]] of calls posted on [http://www.inesc-id.pt/intranet/bolsas/list_fellowships.php INESC ID's website], with both open and closed calls.<br />
* [[Older Calls]]</div>Ramonhttps://www.hlt.inesc-id.pt/wiki/index.php?title=Opportunities_(all_announcements)&diff=7512Opportunities (all announcements)2015-02-23T16:09:35Z<p>Ramon: </p>
<hr />
<div>__NOTOC__<br />
[http://www.inesc-id.pt/ INESC ID] is one of the largest and most dynamic research institutes in Portugal in the areas of telecommunications, information systems and electronics with the status of associated laboratory given by the [http://www.portugal.gov.pt/ Portuguese government]. <br />
<br />
The Laboratorio de Lengua Falada (L2F), or Spoken Language Systems Lab in English, provides an interdisciplinary and creative work environment active in multiple areas of speech and language research. This includes speech recognition, synthesis, and understanding, audio and music processing, multimodal dialogue systems, language and dialect identification, speech-to-speech machine translation, natural language generation, syntactic and semantic processing, summarization, named entity retrieval, question answering and more.<br />
<br />
We also have a powerful IT architecture including robust and fast shared filesystems (AFS, PbFS) for intensive computing operations over large datasets, a computer cluster (with Hadoop, condor) specialized computing units (GPUs, intel-phi) as well as VPN access. <br />
<br />
<br />
<!-- Ramon: I commented this because it did not update --><br />
<!-- <inesc-id what='calls' cc='IP02'></inesc-id> --><br />
<!-- <inesc-id what='calls' cc='IP02' showclosed='yes'></inesc-id>--><br />
<!-- PLEASE POST NEW POSITION ANNOUNCEMENTS THROUGH www.inesc-id.pt --><br />
<br />
== Ph.D. Program CMU-Portugal ==<br />
<br />
* [[Ph.D. Program Carnegie Mellon - PORTUGAL in the area of Language and Information Technologies|'''Ph.D. Program Carnegie Mellon-Portugal''' in the area of Language and Information Technologies]]<br />
<br />
== Post-Doc Fellowship in Project INSIDE ==<br />
<br />
This is a 3 year position, starting in April 2015.<br />
<br />
INSIDE is a joint research project involving Portuguese academic/industrial partners and Carnegie Mellon University. The project explores symbiotic child-robot interactions in joint cooperative activities, with application to the treatment of children with Autism Spectrum Disorder impairments.<br />
<br />
Your role would be related to speech processing in environmentally challenging conditions and with limited adaptation data available. Research will be focused on static and dynamic adaptation of models starting from our current research on deep learning and observation uncertainty techniques.<br />
<br />
We are looking for candidates with following experience:<br />
<br />
* Robust ASR/KWS<br />
* Neural networks<br />
* Python, bash scripting and C/C++<br />
<br />
Additional experience welcome<br />
<br />
* Computational Paralinguistics<br />
* Python-theano<br />
* Knowledge of the Portuguese language<br />
* Missing features / Observation uncertainty<br />
<br />
If interested, send a curriculum vitae (highlighting publications relevant to the position) and a motivation letter to ramon.astudillo@l2f.inesc-id.pt before March 15.<br />
<br />
<br />
<br />
== Opportunities for Final Year Students ==<br />
<br />
(currently closed)<br />
<br />
== Other Information ==<br />
<br />
* A [[Positions|complete list]] of calls posted on [http://www.inesc-id.pt/intranet/bolsas/list_fellowships.php INESC ID's website], with both open and closed calls.<br />
* [[Older Calls]]</div>Ramonhttps://www.hlt.inesc-id.pt/wiki/index.php?title=Opportunities_(all_announcements)&diff=7511Opportunities (all announcements)2015-02-23T16:01:17Z<p>Ramon: </p>
<hr />
<div>__NOTOC__<br />
[http://www.inesc-id.pt/ INESC ID] is one of the largest and most dynamic research institutes in Portugal in the areas of telecommunications, information systems and electronics with the status of associated laboratory given by the [http://www.portugal.gov.pt/ Portuguese government]. <br />
<br />
The Laboratorio de Lengua Falada (L2F), or Spoken Language Systems Lab in English, provides an interdisciplinary and creative work environment active in multiple areas of speech and language research. This includes speech recognition, synthesis, and understanding, audio and music processing, multimodal dialogue systems, language and dialect identification, speech-to-speech machine translation, natural language generation, syntactic and semantic processing, summarization, named entity retrieval, question answering and more.<br />
<br />
<br />
<br />
<!-- Ramon: I commented this because it did not update --><br />
<!-- <inesc-id what='calls' cc='IP02'></inesc-id> --><br />
<!-- <inesc-id what='calls' cc='IP02' showclosed='yes'></inesc-id>--><br />
<!-- PLEASE POST NEW POSITION ANNOUNCEMENTS THROUGH www.inesc-id.pt --><br />
<br />
== Ph.D. Program CMU-Portugal ==<br />
<br />
* [[Ph.D. Program Carnegie Mellon - PORTUGAL in the area of Language and Information Technologies|'''Ph.D. Program Carnegie Mellon-Portugal''' in the area of Language and Information Technologies]]<br />
<br />
== Post-Doc Fellowship in Project INSIDE ==<br />
<br />
This is a 3 year position, starting in April 2015.<br />
<br />
INSIDE is a joint research project involving Portuguese academic/industrial partners and Carnegie Mellon University. The project explores symbiotic child-robot interactions in joint cooperative activities, with application to the treatment of children with Autism Spectrum Disorder impairments.<br />
<br />
Your role would be related to speech processing in environmentally challenging conditions and with limited adaptation data available. Research will be focused on static and dynamic adaptation of models starting from our current research on deep learning and observation uncertainty techniques.<br />
<br />
We are looking for candidates with following experience:<br />
<br />
* Robust ASR/KWS<br />
* Neural networks<br />
* Python, bash scripting and C/C++<br />
<br />
Additional experience welcome<br />
<br />
* Computational Paralinguistics<br />
* Python-theano<br />
* Knowledge of the Portuguese language<br />
* Missing features / Observation uncertainty<br />
<br />
If interested, send a curriculum vitae (highlighting publications relevant to the position) and a motivation letter to ramon.astudillo@l2f.inesc-id.pt before March 15.<br />
<br />
<br />
<br />
== Opportunities for Final Year Students ==<br />
<br />
(currently closed)<br />
<br />
== Other Information ==<br />
<br />
* A [[Positions|complete list]] of calls posted on [http://www.inesc-id.pt/intranet/bolsas/list_fellowships.php INESC ID's website], with both open and closed calls.<br />
* [[Older Calls]]</div>Ramonhttps://www.hlt.inesc-id.pt/wiki/index.php?title=Opportunities_(all_announcements)&diff=7510Opportunities (all announcements)2015-02-23T15:41:45Z<p>Ramon: </p>
<hr />
<div>__NOTOC__<br />
[http://www.inesc-id.pt/ INESC ID] is one of the largest and most dynamic research institutes in Portugal in the areas of telecommunications, information systems and electronics with the status of associated laboratory given by the [http://www.portugal.gov.pt/ Portuguese government]. INESC ID seeks highly qualified candidates holding a PhD with a proven track record and ability to perform independent research.<br />
<br />
<!-- Ramon: I commented this because it did not update --><br />
<!-- <inesc-id what='calls' cc='IP02'></inesc-id> --><br />
<!-- <inesc-id what='calls' cc='IP02' showclosed='yes'></inesc-id>--><br />
<!-- PLEASE POST NEW POSITION ANNOUNCEMENTS THROUGH www.inesc-id.pt --><br />
<br />
== Ph.D. Program CMU-Portugal ==<br />
<br />
* [[Ph.D. Program Carnegie Mellon - PORTUGAL in the area of Language and Information Technologies|'''Ph.D. Program Carnegie Mellon-Portugal''' in the area of Language and Information Technologies]]<br />
<br />
== Post-Doc Fellowship in Project INSIDE ==<br />
<br />
This is a 3 year position, starting in April 2015.<br />
<br />
INSIDE is a joint research project involving Portuguese academic/industrial partners and Carnegie Mellon University. The project explores symbiotic child-robot interactions in joint cooperative activities, with application to the treatment of children with Autism Spectrum Disorder impairments.<br />
<br />
Your role would be related to speech processing in environmentally challenging conditions and with limited adaptation data available. Research will be focused on static and dynamic adaptation of models starting from our current research on deep learning and observation uncertainty techniques.<br />
<br />
We are looking for candidates with following experience:<br />
<br />
* Robust ASR/KWS<br />
* Neural networks<br />
* Python, bash scripting and C/C++<br />
<br />
Additional experience welcome<br />
<br />
* Computational Paralinguistics<br />
* Python-theano<br />
* Knowledge of the Portuguese language<br />
* Missing features / Observation uncertainty<br />
<br />
If interested, send a curriculum vitae (highlighting publications relevant to the position) and a motivation letter to ramon.astudillo@l2f.inesc-id.pt before March 15.<br />
<br />
<br />
<br />
== Opportunities for Final Year Students ==<br />
<br />
(currently closed)<br />
<br />
== Other Information ==<br />
<br />
* A [[Positions|complete list]] of calls posted on [http://www.inesc-id.pt/intranet/bolsas/list_fellowships.php INESC ID's website], with both open and closed calls.<br />
* [[Older Calls]]</div>Ramonhttps://www.hlt.inesc-id.pt/wiki/index.php?title=Opportunities_(all_announcements)&diff=7509Opportunities (all announcements)2015-02-23T15:30:57Z<p>Ramon: </p>
<hr />
<div>__NOTOC__<br />
[http://www.inesc-id.pt/ INESC ID] is one of the largest and most dynamic research institutes in Portugal in the areas of telecommunications, information systems and electronics with the status of associated laboratory given by the [http://www.portugal.gov.pt/ Portuguese government]. INESC ID seeks highly qualified candidates holding a PhD with a proven track record and ability to perform independent research.<br />
<br />
<inesc-id what='calls' cc='IP02'></inesc-id><br />
<!-- <inesc-id what='calls' cc='IP02' showclosed='yes'></inesc-id>--><br />
<!-- PLEASE POST NEW POSITION ANNOUNCEMENTS THROUGH www.inesc-id.pt --><br />
<br />
== Ph.D. Program CMU-Portugal ==<br />
<br />
* [[Ph.D. Program Carnegie Mellon - PORTUGAL in the area of Language and Information Technologies|'''Ph.D. Program Carnegie Mellon-Portugal''' in the area of Language and Information Technologies]]<br />
<br />
== Opportunities for Postgraduate Students (PhDs) ==<br />
<br />
* Post-Doc Fellowship in Project INSIDE [http://www.inesc-id.pt/intranet/bolsas/list_fellowships.php see]<br />
<br />
== Opportunities for Final Year Students ==<br />
<br />
(currently closed)<br />
<br />
== Other Information ==<br />
<br />
* A [[Positions|complete list]] of calls posted on [http://www.inesc-id.pt/intranet/bolsas/list_fellowships.php INESC ID's website], with both open and closed calls.<br />
* [[Older Calls]]</div>Ramonhttps://www.hlt.inesc-id.pt/wiki/index.php?title=Opportunities_(all_announcements)&diff=7508Opportunities (all announcements)2015-02-23T15:29:27Z<p>Ramon: </p>
<hr />
<div>__NOTOC__<br />
[http://www.inesc-id.pt/ INESC ID] is one of the largest and most dynamic research institutes in Portugal in the areas of telecommunications, information systems and electronics with the status of associated laboratory given by the [http://www.portugal.gov.pt/ Portuguese government]. INESC ID seeks highly qualified candidates holding a PhD with a proven track record and ability to perform independent research.<br />
<br />
<inesc-id what='calls' cc='IP02'></inesc-id><br />
<!-- <inesc-id what='calls' cc='IP02' showclosed='yes'></inesc-id> --><br />
<!-- PLEASE POST NEW POSITION ANNOUNCEMENTS THROUGH www.inesc-id.pt --><br />
<br />
== Ph.D. Program CMU-Portugal ==<br />
<br />
* [[Ph.D. Program Carnegie Mellon - PORTUGAL in the area of Language and Information Technologies|'''Ph.D. Program Carnegie Mellon-Portugal''' in the area of Language and Information Technologies]]<br />
<br />
== Opportunities for Postgraduate Students (PhDs) ==<br />
<br />
* Post-Doc Fellowship in Project INSIDE [http://www.inesc-id.pt/intranet/bolsas/list_fellowships.php see]<br />
<br />
== Opportunities for Final Year Students ==<br />
<br />
(currently closed)<br />
<br />
== Other Information ==<br />
<br />
* A [[Positions|complete list]] of calls posted on [http://www.inesc-id.pt/intranet/bolsas/list_fellowships.php INESC ID's website], with both open and closed calls.<br />
* [[Older Calls]]</div>Ramonhttps://www.hlt.inesc-id.pt/wiki/index.php?title=Opportunities_(all_announcements)&diff=7507Opportunities (all announcements)2015-02-23T14:56:50Z<p>Ramon: </p>
<hr />
<div>__NOTOC__<br />
[http://www.inesc-id.pt/ INESC ID] is one of the largest and most dynamic research institutes in Portugal in the areas of telecommunications, information systems and electronics with the status of associated laboratory given by the [http://www.portugal.gov.pt/ Portuguese government]. INESC ID seeks highly qualified candidates holding a PhD with a proven track record and ability to perform independent research.<br />
<br />
<inesc-id what='calls' cc='IP02'></inesc-id><br />
<!-- <inesc-id what='calls' cc='IP02' showclosed='yes'></inesc-id> --><br />
<!-- PLEASE POST NEW POSITION ANNOUNCEMENTS THROUGH www.inesc-id.pt --><br />
<br />
== Ph.D. Program CMU-Portugal ==<br />
<br />
* [[Ph.D. Program Carnegie Mellon - PORTUGAL in the area of Language and Information Technologies|'''Ph.D. Program Carnegie Mellon-Portugal''' in the area of Language and Information Technologies]]<br />
<br />
== Opportunities for Postgraduate Students (PhDs) ==<br />
<br />
(currently closed)<br />
<br />
== Opportunities for Final Year Students ==<br />
<br />
(currently closed)<br />
<br />
== Other Information ==<br />
<br />
* A [[Positions|complete list]] of calls posted on [http://www.inesc-id.pt/intranet/bolsas/list_fellowships.php INESC ID's website], with both open and closed calls.<br />
* [[Older Calls]]</div>Ramonhttps://www.hlt.inesc-id.pt/wiki/index.php?title=Opportunities_(all_announcements)&diff=7506Opportunities (all announcements)2015-02-23T14:56:22Z<p>Ramon: </p>
<hr />
<div>__NOTOC__<br />
[http://www.inesc-id.pt/ INESC ID] is one of the largest and most dynamic research institutes in Portugal in the areas of telecommunications, information systems and electronics with the status of associated laboratory given by the [http://www.portugal.gov.pt/ Portuguese government]. INESC ID seeks highly qualified candidates holding a PhD with a proven track record and ability to perform independent research.<br />
<br />
<inesc-id what='calls' cc='IP02'></inesc-id><br />
<!-- <inesc-id what='calls' cc='IP02' showclosed='yes'></inesc-id> --><br />
<!-- PLEASE POST NEW POSITION ANNOUNCEMENTS THROUGH www.inesc-id.pt --><br />
<br />
== Ph.D. Program CMU-Portugal ==<br />
<br />
* [[Ph.D. Program Carnegie Mellon - PORTUGAL in the area of Language and Information Technologies|'''Ph.D. Program Carnegie Mellon-Portugal''' in the area of Language and Information Technologies]]<br />
<br />
== Opportunities for Postgraduate Students (PhDs) ==<br />
<br />
* [[Opportunities]]<br />
<br />
== Opportunities for Final Year Students ==<br />
<br />
(currently closed)<br />
<br />
== Other Information ==<br />
<br />
* A [[Positions|complete list]] of calls posted on [http://www.inesc-id.pt/intranet/bolsas/list_fellowships.php INESC ID's website], with both open and closed calls.<br />
* [[Older Calls]]</div>Ramonhttps://www.hlt.inesc-id.pt/wiki/index.php?title=Master_Theses_Themes_at_L2F&diff=7505Master Theses Themes at L2F2015-02-23T14:18:38Z<p>Ramon: </p>
<hr />
<div>* Aferição de técnicas de apresentação (Isabel Trancoso, Joaquim Jorge)<br />
<br />
* My-ROBOT: a robotic dialogue based learning companion for a maths game (Ana Paiva, Osabel Trancoso)<br />
<br />
* Simplificação/tradução automática de textos (Fernando Batista, Isabel Trancoso)<br />
<br />
* Ferramenta de pesquisa em áudio para advogados (Alberto Abad, Isabel Trancoso)<br />
<br />
* Microphone network selection and calibration approaches in multi-room environments (Alberto Abad, Aluno: Miguel Matos)<br />
<br />
* Mobile alert: combining human mobile motion detection and voice analysis (Paulo Lobato Correia, Isabel Trancoso)<br />
<br />
* Os arquivos de áudio de ontem e de hoje (Thomas Pellegrini, Isabel Trancoso)<br />
<br />
* Query-by-example speaker search in large data speech collections (Alberto Abad, Isabel Trancoso)<br />
<br />
* Spoken term detection in speech collections using spoken queries (Alberto Abad, Isabel Trancoso)<br />
<br />
* Towards a universal language recognition system (Alberto Abad, Isabel Trancoso)<br />
<br />
* Ensinando um chatbot a responder de modo credível com base em diferentes tipos de feedback (Luísa Coheur)<br />
<br />
*Just.Chat – respondendo como um humano com base em legendas de filmes (Luísa Coheur)<br />
<br />
* SERTO – Sobrevivendo a Erros de Reconhecimento, de Teclado e Outros (Luísa Coheur)<br />
<br />
* POE: PlatafOrma de Escrita (Luísa Coheur)<br />
<br />
* Dizes-me onde vive a Scarlett Johansson? – desenvolvimento de um sistema de pergunta/resposta para português (Luísa Coheur)<br />
<br />
* Medicine.Ask: An intelligent search facility for medicine information (Luísa Coheur)<br />
<br />
* Teach me! Interface verbal para ensino de agentes artificiais<br />
<br />
* Aplicação de métodos de aprendizagem ativa em Língua Natural (Nuno Mamede, Aluno: Nuno Aniceto)<br />
<br />
* Classificador de textos para o ensino de português como segunda língua (Nuno Mamede, Aluno: Pedro Curto)<br />
<br />
* Corretor ortográfico e identificação de sufixos com Transdutores (Nuno Mamede, Aluno: Marco Ferreira)<br />
<br />
* Desambiguação semântica de nomes (Nuno Mamede, Aluno: Rita Policarpo)<br />
<br />
* Desambiguação sintatica/semântica de verbos (Nuno Mamede, Aluno: Gonçalo Suissas)<br />
<br />
* Robust Sumarization of Automatic Transcriptions (David Martins de Matos, Ramón Astudillo)<br />
<br />
* Applications of Music Summarization (David Martins de Matos, Ricardo Ribeiro)<br />
<br />
* Audio Cover Song Identification (David Martins de Matos)<br />
<br />
* Compilation of CUDA-like Kernels for Stream Processing Architectures (advisors: David Martins de Matos, Pedro Tomás; student: Pedro Moreira)<br />
<br />
* Fuzzy Music Recommendation (David Martins de Matos, João Paulo Carvalho)<br />
<br />
* Grid Computing in Organizational Environments (David Martins de Matos, Nuno Roma)<br />
<br />
* Implementation of OpenMP on Application-Specific Architectures (David Martins de Matos, Nuno Roma; student: Cláudia Henriques)<br />
<br />
* Interpretação de Postura e Gestos para Interacção com Robots (David Martins de Matos)<br />
<br />
* Semantics for Big Data in the Music Domain (Helena Sofia Pinto, David Martins de Matos; student: Bruno Dias)<br />
<br />
* Video Summaries for Broadcast News (David Martins de Matos, Ricardo Ribeiro)</div>Ramonhttps://www.hlt.inesc-id.pt/wiki/index.php?title=Opportunities_(all_announcements)&diff=7504Opportunities (all announcements)2015-02-23T14:16:03Z<p>Ramon: </p>
<hr />
<div>__NOTOC__<br />
== Research Staff Positions ==<br />
<br />
[http://www.inesc-id.pt/ INESC ID] is one of the largest and most dynamic research institutes in Portugal in the areas of telecommunications, information systems and electronics with the status of associated laboratory given by the [http://www.portugal.gov.pt/ Portuguese government]. INESC ID seeks highly qualified candidates holding a PhD with a proven track record and ability to perform independent research.<br />
<br />
In the Spoken Language Systems area, we are accepting [http://www.inesc-id.pt/intranet/bolsas/list_fellowships.php#37 applications] for researchers in the following lines:<br />
<br />
# Spoken and multimodal dialog systems<br />
# Speech-to-speech translation<br />
# Semantic processing of multimedia content<br />
<br />
Do You Want to Work with Us? Please send us your CV.<br />
<br />
<inesc-id what='calls' cc='IP02'></inesc-id><br />
<!-- <inesc-id what='calls' cc='IP02' showclosed='yes'></inesc-id> --><br />
<!-- PLEASE POST NEW POSITION ANNOUNCEMENTS THROUGH www.inesc-id.pt --><br />
<br />
== Ph.D. Program CMU-Portugal ==<br />
<br />
* [[Ph.D. Program Carnegie Mellon - PORTUGAL in the area of Language and Information Technologies|'''Ph.D. Program Carnegie Mellon-Portugal''' in the area of Language and Information Technologies]]<br />
<br />
== Opportunities for Postgraduate Students (PhDs) ==<br />
<br />
* [[Opportunities]]<br />
<br />
== Opportunities for Final Year Students ==<br />
<br />
(currently closed)<br />
<br />
== Other Information ==<br />
<br />
* A [[Positions|complete list]] of calls posted on [http://www.inesc-id.pt/intranet/bolsas/list_fellowships.php INESC ID's website], with both open and closed calls.<br />
* [[Older Calls]]</div>Ramonhttps://www.hlt.inesc-id.pt/wiki/index.php?title=Press_Kits&diff=7503Press Kits2015-02-23T14:09:17Z<p>Ramon: </p>
<hr />
<div>THIS PAGE IS DEPRECATED. <br />
<br />
__NOTOC__<br />
* '''[http://www.l2f.inesc-id.pt/presskits/All-L2F-Flyers-v2.pdf Media Brochure]'''<br />
<br />
== About us ==<br />
* [http://www.l2f.inesc-id.pt/presskits/0.1-Flyer-L2F-v1.pdf L2F - Spoken Language Systems Laboratory]<br />
<br />
== Research Areas & Activities ==<br />
* [http://www.l2f.inesc-id.pt/presskits/1.0-Flyer-SPMD-v2.pdf Semantic Processing of Multimedia contents]<br />
* [http://www.l2f.inesc-id.pt/presskits/2.0-Flyer-SMDS-v2.pdf Spoken / Multimodal Dialogue Systems]<br />
* [http://www.l2f.inesc-id.pt/presskits/3.0-Flyer-SMT-v2.pdf Speech-to-Speech Machine Translation]<br />
* [http://www.l2f.inesc-id.pt/presskits/1.1-Flyer-DTBs-v4.pdf Digital Talking Books]<br />
<br />
== Core Technologies ==<br />
* [http://www.l2f.inesc-id.pt/presskits/4.1-Flyer-Audimus-v1.pdf AUDIMUS – Portuguese Speech RecRecognizer]<br />
* [http://www.l2f.inesc-id.pt/presskits/4.2-Flyer-Dixi-v2.pdf DIXI - Portuguese Text-to-Speech Synthesizer]<br />
* [http://www.l2f.inesc-id.pt/presskits/4.3-Flyer-Face-v1.pdf FACE - Synthetic Talking Face]<br />
* [http://www.l2f.inesc-id.pt/presskits/4.4-Flyer-Tools-v1.pdf Natural Language Processing Tools]<br />
<br />
== Ongoing Projects ==<br />
* [https://www.l2f.inesc-id.pt/wiki/index.php/Projects List of Projects]</div>Ramonhttps://www.hlt.inesc-id.pt/wiki/index.php?title=Robust_Speech_Processing_using_Observation_Uncertainty_and_Uncertainty_Propagation&diff=7502Robust Speech Processing using Observation Uncertainty and Uncertainty Propagation2015-02-10T01:21:38Z<p>Ramon: </p>
<hr />
<div>== Call for Papers for a Special Session at Interspeech 2015 *Deadline March 20* ==<br />
<br />
We are pleased to announce a special session at [http://interspeech2015.org/ Interspeech 2015] in Dresden focusing on Observation Uncertainty (OU) and Uncertainty Propagation (UP) research in speech processing.<br />
<br />
Observation Uncertainty (OU) and Uncertainty Propagation (UP) techniques for robust speech processing have shown notable success across various sub-domains as e.g. ASR or Speaker Recognition. Despite its healthy status, OU/UP research in speech processing is currently fragmented across various disciplines and application domains. Consequently, research papers on uncertainty have to compete with other techniques in e.g. ASR or Speaker Identification tracks. This generally discourages the proposal of novel or bold ideas pertaining to the topic. This is particularly relevant in the current environment in which the strong technology transfer and real-data-rich problems force disciplines to adapt themselves more quickly. In particular, although investigations on how to port OU/UP techniques to Deep Neural Networks (DNNs) exist, this area is still yet to be studied. Given the disruptive character of DNNs and the fact that they are not amenable to classic adaptation techniques, this poses a very interesting application dimension for OU/UP techniques. For these reasons, we have proposed a special session devoted to uncertainty as an opportunity for researchers of different sub-domains of speech processing to share ideas, explore new application domains as e.g. paralinguistics, or text-based uncertainties and consolidate OU/UP research in DNNs. <br />
<br />
The session would cover <br />
<br />
* Methods exploiting observation and parameter uncertainty as well as uncertainty propagation for robust speech processing e.g., for inference, learning or adaptation<br />
<br />
* We specially welcome papers on the topic deep learning and uncertainty. We are confident this is the right moment to tackle this issue and already expect works on the topic<br />
<br />
* Uncertainty in training data: “BIG” data is upon us, but the quality of that ‘found’ data is not the same as carefully collected data. What are the best ways to address data uncertainty when ground-truth is unknown? <br />
<br />
* All sub-fields are welcome, since this is not restricted to only ASR and speaker recognition communities but also works in e.g., paralinguistics or transcription level uncertainties (lattices).<br />
<br />
We would also like to use this opportunity to motivate the sharing of software tools based on uncertainty techniques among the community as well as reproducible research. For this purpose we will provide software for uncertainty-related techniques, such as tools for the estimation of observation uncertainties and for their application to ASR, through the wiki page of the Robust Speech Processing Special Interest Group of ISCA (RoSP-SIG).<br />
<br />
https://wiki.inria.fr/rosp/Software<br />
<br />
All updates on the session progress and available tools will be reported on this website.<br />
<br />
Update 2015-02-10: Tools for uncertainty propagation and observation uncertainty using HTK/Kaldi released (see WiKi link above). <br />
<br />
Do not hesitate to contact us if you have any question. <br />
<br />
Ramón Fernandez Astudillo (ramon.astudillo@inesc-id.pt)<br />
Shinji Watanabe (watanabe@merl.com)<br />
Ahmed Hussen AbdelAziz (Ahmed.HussenAbdelAziz@rub.de)<br />
Dorothea Kolossa (dorothea.kolossa@rub.de)</div>Ramonhttps://www.hlt.inesc-id.pt/wiki/index.php?title=Robust_Speech_Processing_using_Observation_Uncertainty_and_Uncertainty_Propagation&diff=7492Robust Speech Processing using Observation Uncertainty and Uncertainty Propagation2015-01-20T13:43:56Z<p>Ramon: </p>
<hr />
<div>== Call for Papers for a Special Session at Interspeech 2015 *Deadline March 20* ==<br />
<br />
We are pleased to announce a special session at [http://interspeech2015.org/ Interspeech 2015] in Dresden focusing on Observation Uncertainty (OU) and Uncertainty Propagation (UP) research in speech processing.<br />
<br />
Observation Uncertainty (OU) and Uncertainty Propagation (UP) techniques for robust speech processing have shown notable success across various sub-domains as e.g. ASR or Speaker Recognition. Despite its healthy status, OU/UP research in speech processing is currently fragmented across various disciplines and application domains. Consequently, research papers on uncertainty have to compete with other techniques in e.g. ASR or Speaker Identification tracks. This generally discourages the proposal of novel or bold ideas pertaining to the topic. This is particularly relevant in the current environment in which the strong technology transfer and real-data-rich problems force disciplines to adapt themselves more quickly. In particular, although investigations on how to port OU/UP techniques to Deep Neural Networks (DNNs) exist, this area is still yet to be studied. Given the disruptive character of DNNs and the fact that they are not amenable to classic adaptation techniques, this poses a very interesting application dimension for OU/UP techniques. For these reasons, we have proposed a special session devoted to uncertainty as an opportunity for researchers of different sub-domains of speech processing to share ideas, explore new application domains as e.g. paralinguistics, or text-based uncertainties and consolidate OU/UP research in DNNs. <br />
<br />
The session would cover <br />
<br />
* Methods exploiting observation and parameter uncertainty as well as uncertainty propagation for robust speech processing e.g., for inference, learning or adaptation<br />
<br />
* We specially welcome papers on the topic deep learning and uncertainty. We are confident this is the right moment to tackle this issue and already expect works on the topic<br />
<br />
* Uncertainty in training data: “BIG” data is upon us, but the quality of that ‘found’ data is not the same as carefully collected data. What are the best ways to address data uncertainty when ground-truth is unknown? <br />
<br />
* All sub-fields are welcome, since this is not restricted to only ASR and speaker recognition communities but also works in e.g., paralinguistics or transcription level uncertainties (lattices).<br />
<br />
We would also like to use this opportunity to motivate the sharing of software tools based on uncertainty techniques among the community as well as reproducible research. For this purpose we will provide software for uncertainty-related techniques, such as tools for the estimation of observation uncertainties and for their application to ASR, through the wiki page of the Robust Speech Processing Special Interest Group of ISCA (RoSP-SIG).<br />
<br />
https://wiki.inria.fr/rosp/Software<br />
<br />
All updates on the session progress and available tools will be reported on this website.<br />
<br />
Do not hesitate to contact us if you have any question. <br />
<br />
Ramón Fernandez Astudillo (ramon.astudillo@inesc-id.pt)<br />
Shinji Watanabe (watanabe@merl.com)<br />
Ahmed Hussen AbdelAziz (Ahmed.HussenAbdelAziz@rub.de)<br />
Dorothea Kolossa (dorothea.kolossa@rub.de)</div>Ramonhttps://www.hlt.inesc-id.pt/wiki/index.php?title=Robust_Speech_Processing_using_Observation_Uncertainty_and_Uncertainty_Propagation&diff=7486Robust Speech Processing using Observation Uncertainty and Uncertainty Propagation2015-01-14T08:06:22Z<p>Ramon: </p>
<hr />
<div>== Call for Papers for a Special Session at Interspeech 2015 *Deadline March 20* ==<br />
<br />
We are pleased to announce a special session at [http://interspeech2015.org/ Interspeech 2015] in Dresden focusing on Observation Uncertainty (OU) and Uncertainty Propagation (UP) research in speech processing.<br />
<br />
Observation Uncertainty (OU) and Uncertainty Propagation (UP) techniques for robust speech processing have shown notable success across various sub-domains as e.g. ASR or Speaker Recognition. Despite its healthy status, OU/UP research in speech processing is currently fragmented across various disciplines and application domains. Consequently, research papers on uncertainty have to compete with other techniques in e.g. ASR or Speaker Identification tracks. This generally discourages the proposal of novel or bold ideas pertaining to the topic. This is particularly relevant in the current environment in which the strong technology transfer and real-data-rich problems force disciplines to adapt themselves more quickly. In particular, although investigations on how to port OU/UP techniques to Deep Neural Networks (DNNs) exist, this area is still yet to be studied. Given the disruptive character of DNNs and the fact that they are not amenable to classic adaptation techniques, this poses a very interesting application dimension for OU/UP techniques. For these reasons, we have proposed a special session devoted to uncertainty as an opportunity for researchers of different sub-domains of speech processing to share ideas, explore new application domains as e.g. paralinguistics, or text-based uncertainties and consolidate OU/UP research in DNNs. <br />
<br />
The session would cover <br />
<br />
* Methods exploiting observation and parameter uncertainty as well as uncertainty propagation for robust speech processing e.g., for interference, learning or adaptation<br />
<br />
* We specially welcome papers on the topic deep learning and uncertainty. We are confident this is the right moment to tackle this issue and already expect works on the topic<br />
<br />
* Uncertainty in training data: “BIG” data is upon us, but the quality of that ‘found’ data is not the same as carefully collected data. What are the best ways to address data uncertainty when ground-truth is unknown? <br />
<br />
* All sub-fields are welcome, since this is not restricted to only ASR and speaker recognition communities but also works in e.g., paralinguistics or transcription level uncertainties (lattices).<br />
<br />
We would also like to use this opportunity to motivate the sharing of software tools based on uncertainty techniques among the community as well as reproducible research. For this purpose we will provide software for uncertainty-related techniques, such as tools for the estimation of observation uncertainties and for their application to ASR, through the wiki page of the Robust Speech Processing Special Interest Group of ISCA (RoSP-SIG).<br />
<br />
https://wiki.inria.fr/rosp/Software<br />
<br />
All updates on the session progress and available tools will be reported on this website.<br />
<br />
Do not hesitate to contact us if you have any question. <br />
<br />
Ramón Fernandez Astudillo (ramon.astudillo@inesc-id.pt)<br />
Shinji Watanabe (watanabe@merl.com)<br />
Ahmed Hussen AbdelAziz (Ahmed.HussenAbdelAziz@rub.de)<br />
Dorothea Kolossa (dorothea.kolossa@rub.de)</div>Ramonhttps://www.hlt.inesc-id.pt/wiki/index.php?title=Robust_Speech_Processing_using_Observation_Uncertainty_and_Uncertainty_Propagation&diff=7485Robust Speech Processing using Observation Uncertainty and Uncertainty Propagation2015-01-13T14:36:59Z<p>Ramon: </p>
<hr />
<div>== Call for Papers for a Special Session at Interspeech 2015 *Deadline March 20* ==<br />
<br />
We are pleased to announce a special session at [http://interspeech2015.org/ Interspeech 2015] in Dresden focusing on Observation Uncertainty (OU) and Uncertainty Propagation (UP) research in speech processing.<br />
<br />
Observation Uncertainty (OU) and Uncertainty Propagation (UP) techniques for robust speech processing have shown notable success across various sub-domains as e.g. ASR or Speaker Recognition. Despite its healthy status, OU/UP research in speech processing is currently fragmented across various disciplines and application domains. Consequently, research papers on uncertainty have to compete with other techniques in e.g. ASR or Speaker Identification tracks. This generally discourages the proposal of novel or bold ideas pertaining to the topic. This is particularly relevant in the current environment in which the strong technology transfer and real-data-rich problems force disciplines to adapt themselves more quickly. In particular, although investigations on how to port OU/UP techniques to Deep Neural Networks (DNNs) exist, this area is still yet to be studied. Given the disruptive character of DNNs and the fact that they are not amenable to classic adaptation techniques, this poses a very interesting application dimension for OU/UP techniques. For these reasons, we have proposed a special session devoted to uncertainty as an opportunity for researchers of different sub-domains of speech processing to share ideas, explore new application domains as e.g. paralinguistics, or text-based uncertainties and consolidate OU/UP research in DNNs. <br />
<br />
The session would cover <br />
<br />
* Methods exploiting observation and parameter uncertainty as well as uncertainty propagation for robust speech processing (e.g., for interference, learning or adaptation)<br />
<br />
* We specially welcome papers on the topic deep learning and uncertainty. We are confident this is the right moment to tackle this issue and already expect works on the topic<br />
<br />
* Uncertainty in training data: “BIG” data is upon us, but the quality of that ‘found’ data is not the same as carefully collected data. What are the best ways to address data uncertainty when ground-truth is unknown? <br />
<br />
* All sub-fields are welcome, since this is not restricted to only ASR and speaker recognition communities, but also exploratory speech/language domains (e.g., paralinguistics or transcription level uncertainties (lattices)).<br />
<br />
In order to motivate exploratory ideas and facilitate research we will provide tools for various uncertain feature extractions as well as dynamic compensation code for HTK. Their use will not be obligatory. They will just provided for those interested.<br />
<br />
All updates on the session progress and available tools will be reported on this website.<br />
<br />
Do not hesitate to contact us if you have any question. <br />
<br />
Ramón Fernandez Astudillo (ramon.astudillo@inesc-id.pt)<br />
Shinji Watanabe (watanabe@merl.com)<br />
Ahmed Hussen AbdelAziz (Ahmed.HussenAbdelAziz@rub.de)<br />
Dorothea Kolossa (dorothea.kolossa@rub.de)</div>Ramonhttps://www.hlt.inesc-id.pt/wiki/index.php?title=Robust_Speech_Processing_using_Observation_Uncertainty_and_Uncertainty_Propagation&diff=7466Robust Speech Processing using Observation Uncertainty and Uncertainty Propagation2015-01-07T19:13:46Z<p>Ramon: </p>
<hr />
<div>== Call for Papers for a Special Session at Interspeech 2015 *Deadline March 20* ==<br />
<br />
We are pleased to announce a special session at [http://interspeech2015.org/ Interspeech 2015] in Dresden focusing on Observation Uncertainty (OU) and Uncertainty Propagation (UP) research in speech processing.<br />
<br />
Observation Uncertainty (OU) and Uncertainty Propagation (UP) techniques for robust speech processing have shown notable success across various sub-domains as e.g. ASR or Speaker Recognition. Despite its healthy status, OU/UP research in speech processing is currently fragmented across various disciplines and application domains. Consequently, research papers on uncertainty have to compete with other techniques in e.g. ASR or Speaker Identification tracks. This generally discourages the proposal of novel or bold ideas pertaining to the topic. This is particularly relevant in the current environment in which the strong technology transfer and real-data-rich problems force disciplines to adapt themselves more quickly. In particular, although investigations on how to port OU/UP techniques to Deep Neural Networks (DNNs) exist, this area is still yet to be studied. Given the disruptive character of DNNs and the fact that they are not amenable to classic adaptation techniques, this poses a very interesting application dimension for OU/UP techniques. For these reasons, we have proposed a special session devoted to uncertainty as an opportunity for researchers of different sub-domains of speech processing to share ideas, explore new application domains as e.g. paralinguistics, or text-based uncertainties and consolidate OU/UP research in DNNs. <br />
<br />
The session would cover <br />
<br />
* Methods exploiting observation and parameter uncertainty as well as uncertainty propagation for robust speech processing e.g. for inference, learning or adaptation<br />
<br />
* Especially welcome are papers on the topic deep learning and uncertainty. We are confident this is the right moment to tackle this issue and already expect works on the topic<br />
<br />
* All sub-fields are wellcome. Not only ASR and speaker recognition communities but also exploratory works on e.g. paralinguistics or transcription level uncertainties (lattices).<br />
<br />
In order to motivate exploratory ideas and facilitate research we will provide tools for various uncertain feature extractions as well as dynamic compensation code for HTK and Kaldi. Their use will not be obligatory. They will just provided for those interested.<br />
<br />
Do not hesitate to contact us if you have any question. All updates on the session progress and available tools will be reported on this website.<br />
<br />
Ramón Fernandez Astudillo (ramon.astudillo@inesc-id.pt)<br />
<br />
Shinji Watanabe (watanabe@merl.com)<br />
<br />
Ahmed Hussen AbdelAziz (Ahmed.HussenAbdelAziz@rub.de)<br />
<br />
Dorothea Kolossa (dorothea.kolossa@rub.de)</div>Ramonhttps://www.hlt.inesc-id.pt/wiki/index.php?title=Robust_Speech_Processing_using_Observation_Uncertainty_and_Uncertainty_Propagation&diff=7465Robust Speech Processing using Observation Uncertainty and Uncertainty Propagation2015-01-07T18:27:24Z<p>Ramon: </p>
<hr />
<div>== Call for Papers for a Special Session at Interspeech 2015 *Deadline March 20* ==<br />
<br />
We are pleased to announce a special session at [http://interspeech2015.org/ Interspeech 2015] in Dresden focusing on Observation Uncertainty (OU) and Uncertainty Propagation (UP) research in speech processing.<br />
<br />
Despite its healthy status, OU/UP research in speech processing is currently fragmented across various disciplines and application domains. Consequently, research papers on uncertainty have to compete with other techniques in e.g. ASR or Speaker Identification tracks. This generally discourages the proposal of novel or bold ideas pertaining to the topic. This is particularly relevant in the current environment in which the strong technology transfer and real-data-rich problems force disciplines to adapt themselves more quickly. In particular, although investigations on how to port OU/UP techniques to Deep Neural Networks (DNNs) exist, this area is still yet to be studied. Given the disruptive character of DNNs and the fact that they are not amenable to classic adaptation techniques, this poses a very interesting application dimension for OU/UP techniques. For these reasons, we have proposed a special session devoted to uncertainty as an opportunity for researchers of different sub-domains of speech processing to share ideas, explore new application domains as e.g. paralinguistics, or text-based uncertainties and consolidate OU/UP research in DNNs. <br />
<br />
The session would cover <br />
<br />
* Methods exploiting observation and parameter uncertainty as well as uncertainty propagation for robust speech processing e.g. for inference, learning or adaptation<br />
<br />
* Especially welcome are papers on the topic deep learning and uncertainty. We are confident this is the right moment to tackle this issue and already expect works on the topic<br />
<br />
* All sub-fields are wellcome. Not only ASR and speaker recognition communities but also exploratory works on e.g. paralinguistics or transcription level uncertainties (lattices).<br />
<br />
In order to motivate exploratory ideas and facilitate research we will provide tools for various uncertain feature extractions as well as dynamic compensation code for HTK and Kaldi. Their use will not be obligatory. They will just provided for those interested.<br />
<br />
Do not hesitate to contact us if you have any question. All updates on the session progress and available tools will be reported on this website.<br />
<br />
Ramón Fernandez Astudillo (ramon.astudillo@inesc-id.pt)<br />
<br />
Shinji Watanabe (watanabe@merl.com)<br />
<br />
Ahmed Hussen AbdelAziz (Ahmed.HussenAbdelAziz@rub.de)<br />
<br />
Dorothea Kolossa (dorothea.kolossa@rub.de)</div>Ramonhttps://www.hlt.inesc-id.pt/wiki/index.php?title=Robust_Speech_Processing_using_Observation_Uncertainty_and_Uncertainty_Propagation&diff=7464Robust Speech Processing using Observation Uncertainty and Uncertainty Propagation2015-01-07T18:26:45Z<p>Ramon: </p>
<hr />
<div>== Call for Papers for a Special Session at Interspeech 2015 *Deadline March 20* ==<br />
<br />
We are pleased to announce a special session at [http://interspeech2015.org/ Interspeech 2015] in Dresden focusing on Observation Uncertainty (OU) and Uncertainty Propagation (UP) research in speech processing.<br />
<br />
Despite its healthy status, OU/UP research in speech processing is currently fragmented across various disciplines and application domains. Consequently, research papers on uncertainty have to compete with other techniques in e.g. ASR or Speaker Identification tracks. This generally discourages the proposal of novel or bold ideas pertaining to the topic. This is particularly relevant in the current environment in which the strong technology transfer and real-data-rich problems force disciplines to adapt themselves more quickly. In particular, although investigations on how to port OU/UP techniques to Deep Neural Networks (DNNs) exist, this area is still yet to be studied. Given the disruptive character of DNNs and the fact that they are not amenable to classic adaptation techniques, this poses a very interesting application dimension for OU/UP techniques. For these reasons, we have proposed a special session devoted to uncertainty as an opportunity for researchers of different sub-domains of speech processing to share ideas, explore new application domains as e.g. paralinguistics, or text-based uncertainties and consolidate OU/UP research in DNNs. <br />
<br />
The session would cover <br />
<br />
* Methods exploiting observation and parameter uncertainty as well as uncertainty propagation for robust speech processing e.g. for inference, learning or adaptation<br />
<br />
* Especially welcome are papers on the topic deep learning and uncertainty. We are confident this is the right moment to tackle this issue and already expect works on the topic<br />
<br />
* All sub-fields are wellcome. Not only ASR and speaker recognition communities but also exploratory works on e.g. paralinguistics or transcription level uncertainties (lattices).<br />
<br />
In order to motivate exploratory ideas and facilitate research we will provide tools for various uncertain feature extractions as well as dynamic compensation code for HTK and Kaldi. Their use of these tools will *not* obligatory. They will just provided for those interested.<br />
<br />
Do not hesitate to contact us if you have any question. All updates on the session progress and available tools will be reported on this website.<br />
<br />
Ramón Fernandez Astudillo (ramon.astudillo@inesc-id.pt)<br />
<br />
Shinji Watanabe (watanabe@merl.com)<br />
<br />
Ahmed Hussen AbdelAziz (Ahmed.HussenAbdelAziz@rub.de)<br />
<br />
Dorothea Kolossa (dorothea.kolossa@rub.de)</div>Ramonhttps://www.hlt.inesc-id.pt/wiki/index.php?title=Robust_Speech_Processing_using_Observation_Uncertainty_and_Uncertainty_Propagation&diff=7463Robust Speech Processing using Observation Uncertainty and Uncertainty Propagation2015-01-07T17:30:50Z<p>Ramon: </p>
<hr />
<div>== Call for Papers for a Special Session at Interspeech 2015 *Deadline March 20* ==<br />
<br />
We are pleased to announce a special session at [http://interspeech2015.org/ Interspeech 2015] in Dresden focusing on Observation Uncertainty (OU) and Uncertainty Propagation (UP) research in speech processing.<br />
<br />
Despite its healthy status, OU/UP research in speech processing is currently fragmented across various disciplines and application domains. Consequently, research papers on uncertainty have to compete with other techniques in e.g. ASR or Speaker Identification tracks. This generally discourages the proposal of novel or bold ideas pertaining to the topic. This is particularly relevant in the current environment in which the strong technology transfer and real-data-rich problems force disciplines to adapt themselves more quickly. In particular, although investigations on how to port OU/UP techniques to Deep Neural Networks (DNNs) exist, this area is still yet to be studied. Given the disruptive character of DNNs and the fact that they are not amenable to classic adaptation techniques, this poses a very interesting application dimension for OU/UP techniques. For these reasons, we have proposed a special session devoted to uncertainty as an opportunity for researchers of different sub-domains of speech processing to share ideas, explore new application domains as e.g. paralinguistics, or text-based uncertainties and consolidate OU/UP research in DNNs. <br />
<br />
The session would cover <br />
<br />
* Methods exploiting observation and parameter uncertainty as well as uncertainty propagation for robust speech processing e.g. for inference, learning or adaptation<br />
<br />
* Especially welcome are papers on the topic deep learning and uncertainty. We are confident this is the right moment to tackle this issue and already expect works on the topic<br />
<br />
* All sub-fields are wellcome. Not only ASR and speaker recognition communities but also exploratory works on e.g. paralinguistics or transcription level uncertainties (lattices).<br />
<br />
In order to motivate exploratory ideas, and facilitate research we will provide the following tools to those interested in participating:<br />
<br />
* Feature extractors for different types of uncertainties (Wiener posterior, sparsity assumption, Supergaussian) compatible with HTK and Kaldi in Matlab<br />
<br />
* Patches for uncertainty decoding and modified imputation in HTK, thus allowing to use uncertainties in the existing recipes.<br />
<br />
* Patches to read Matlab produced DNN posteriors with Kaldi, thus allowing to implement uncertainty techniques based on DNNs in Matlab. <br />
<br />
The use of these tools is *not* obligatory. They are just provided for those interested.<br />
<br />
Do not hesitate to contact us if you have any question. All updates on the session progress and available tools will be reported on this website.<br />
<br />
Ramón Fernandez Astudillo (ramon.astudillo@inesc-id.pt)<br />
<br />
Shinji Watanabe (watanabe@merl.com)<br />
<br />
Ahmed Hussen AbdelAziz (Ahmed.HussenAbdelAziz@rub.de)<br />
<br />
Dorothea Kolossa (dorothea.kolossa@rub.de)</div>Ramonhttps://www.hlt.inesc-id.pt/wiki/index.php?title=Robust_Speech_Processing_using_Observation_Uncertainty_and_Uncertainty_Propagation&diff=7462Robust Speech Processing using Observation Uncertainty and Uncertainty Propagation2015-01-07T17:29:43Z<p>Ramon: </p>
<hr />
<div>== Call for Papers for a Special Session at Interspeech 2015 *Deadline March 20* ==<br />
<br />
We are pleased to announce a special session at [http://interspeech2015.org/ Interspeech 2015] in Dresden focusing on Observation Uncertainty (OU) and Uncertainty Propagation (UP) research in speech processing.<br />
<br />
Despite its healthy status, OU/UP research in speech processing is currently fragmented across various disciplines and application domains. Consequently, research papers on uncertainty have to compete with other techniques in e.g. ASR or Speaker Identification tracks. This generally discourages the proposal of novel or bold ideas pertaining to the topic. This is particularly relevant in the current environment in which the strong technology transfer and real-data-rich problems force disciplines to adapt themselves more quickly. In particular, although investigations on how to port OU/UP techniques to Deep Neural Networks (DNNs) exist, this area is still yet to be studied. Given the disruptive character of DNNs and the fact that they are not amenable to classic adaptation techniques, this poses a very interesting application dimension for OU/UP techniques. For these reasons, we have proposed a special session devoted to uncertainty as an opportunity for researchers of different sub-domains of speech processing to share ideas, explore new application domains as e.g. paralinguistics, or text based uncertainties and consolidate OU/UP research in DNNs. <br />
<br />
The session would cover <br />
<br />
* Methods exploiting observation and parameter uncertainty as well as uncertainty propagation for robust speech processing e.g. for inference, learning or adaptation<br />
<br />
* Especially welcome are papers on the topic deep learning and uncertainty. We are confident this is the right moment to tackle this issue and already expect works on the topic<br />
<br />
* All sub-fields are wellcome. Not only ASR and speaker recognition communities but also exploratory works on e.g. paralinguistics or transcription level uncertainties (lattices).<br />
<br />
In order to motivate exploratory ideas, and facilitate research we will provide the following tools to those interested in participating:<br />
<br />
* Feature extractors for different types of uncertainties (Wiener posterior, sparsity assumption, Supergaussian) compatible with HTK and Kaldi in Matlab<br />
<br />
* Patches for uncertainty decoding and modified imputation in HTK, thus allowing to use uncertainties in the existing recipes.<br />
<br />
* Patches to read Matlab produced DNN posteriors with Kaldi, thus allowing to implement uncertainty techniques based on DNNs in Matlab. <br />
<br />
The use of these tools is *not* obligatory. They are just provided for those interested.<br />
<br />
Do not hesitate to contact us if you have any question. All updates on the session progress and available tools will be reported on this website.<br />
<br />
Ramón Fernandez Astudillo (ramon.astudillo@inesc-id.pt)<br />
<br />
Shinji Watanabe (watanabe@merl.com)<br />
<br />
Ahmed Hussen AbdelAziz (Ahmed.HussenAbdelAziz@rub.de)<br />
<br />
Dorothea Kolossa (dorothea.kolossa@rub.de)</div>Ramonhttps://www.hlt.inesc-id.pt/wiki/index.php?title=Robust_Speech_Processing_using_Observation_Uncertainty_and_Uncertainty_Propagation&diff=7461Robust Speech Processing using Observation Uncertainty and Uncertainty Propagation2015-01-07T17:29:09Z<p>Ramon: </p>
<hr />
<div>== Call for Papers for a Special Session at Interspeech 2015 *Deadline 20 March* ==<br />
<br />
We are pleased to announce a special session at [http://interspeech2015.org/ Interspeech 2015] in Dresden focusing on Observation Uncertainty (OU) and Uncertainty Propagation (UP) research in speech processing.<br />
<br />
Despite its healthy status, OU/UP research in speech processing is currently fragmented across various disciplines and application domains. Consequently, research papers on uncertainty have to compete with other techniques in e.g. ASR or Speaker Identification tracks. This generally discourages the proposal of novel or bold ideas pertaining to the topic. This is particularly relevant in the current environment in which the strong technology transfer and real-data-rich problems force disciplines to adapt themselves more quickly. In particular, although investigations on how to port OU/UP techniques to Deep Neural Networks (DNNs) exist, this area is still yet to be studied. Given the disruptive character of DNNs and the fact that they are not amenable to classic adaptation techniques, this poses a very interesting application dimension for OU/UP techniques. For these reasons, we have proposed a special session devoted to uncertainty as an opportunity for researchers of different sub-domains of speech processing to share ideas, explore new application domains as e.g. paralinguistics, or text based uncertainties and consolidate OU/UP research in DNNs. <br />
<br />
The session would cover <br />
<br />
* Methods exploiting observation and parameter uncertainty as well as uncertainty propagation for robust speech processing e.g. for inference, learning or adaptation<br />
<br />
* Especially welcome are papers on the topic deep learning and uncertainty. We are confident this is the right moment to tackle this issue and already expect works on the topic<br />
<br />
* All sub-fields are wellcome. Not only ASR and speaker recognition communities but also exploratory works on e.g. paralinguistics or transcription level uncertainties (lattices).<br />
<br />
In order to motivate exploratory ideas, and facilitate research we will provide the following tools to those interested in participating:<br />
<br />
* Feature extractors for different types of uncertainties (Wiener posterior, sparsity assumption, Supergaussian) compatible with HTK and Kaldi in Matlab<br />
<br />
* Patches for uncertainty decoding and modified imputation in HTK, thus allowing to use uncertainties in the existing recipes.<br />
<br />
* Patches to read Matlab produced DNN posteriors with Kaldi, thus allowing to implement uncertainty techniques based on DNNs in Matlab. <br />
<br />
The use of these tools is *not* obligatory. They are just provided for those interested.<br />
<br />
Do not hesitate to contact us if you have any question. All updates on the session progress and available tools will be reported on this website.<br />
<br />
Ramón Fernandez Astudillo (ramon.astudillo@inesc-id.pt)<br />
<br />
Shinji Watanabe (watanabe@merl.com)<br />
<br />
Ahmed Hussen AbdelAziz (Ahmed.HussenAbdelAziz@rub.de)<br />
<br />
Dorothea Kolossa (dorothea.kolossa@rub.de)</div>Ramonhttps://www.hlt.inesc-id.pt/wiki/index.php?title=Robust_Speech_Processing_using_Observation_Uncertainty_and_Uncertainty_Propagation&diff=7460Robust Speech Processing using Observation Uncertainty and Uncertainty Propagation2015-01-07T17:09:21Z<p>Ramon: </p>
<hr />
<div>== Call for Papers for a Special Session at Interspeech 2015 *Deadline 20 March* ==<br />
<br />
We are pleased to announce a special session at [http://interspeech2015.org/ Interspeech 2015] in Dresden. The session will focusi on uncertainty for robust speech processing across all sub-fields. This would cover <br />
<br />
* Methods exploiting observation and parameter uncertainty as well as uncertainty propagation for robust speech processing e.g. for inference, learning or adaptation<br />
<br />
* Especially welcome are papers on the topic deep learning and uncertainty. We are confident this is the right moment to tackle this issue and already expect works on the topic<br />
<br />
* All sub-fields are wellcome. Not only ASR and speaker recognition communities but also exploratory works on e.g. paralinguistics or transcription level uncertainties (lattices).<br />
<br />
In order to motivate exploratory ideas, and facilitate research we will provide the following tools to those interested in participating:<br />
<br />
* Feature extractors for different types of uncertainties (Wiener posterior, sparsity assumption, Supergaussian) compatible with HTK and Kaldi in Matlab<br />
<br />
* Patches for uncertainty decoding and modified imputation in HTK, thus allowing to use uncertainties in the existing recipes.<br />
<br />
* Patches to read Matlab produced DNN posteriors with Kaldi, thus allowing to implement uncertainty techniques based on DNNs in Matlab. <br />
<br />
The use of these tools is *not* obligatory. They are just provided for those interested.<br />
<br />
Do not hesitate to contact us if you have any question<br />
<br />
Ramón Fernandez Astudillo (ramon.astudillo@inesc-id.pt),<br />
Shinji Watanabe (watanabe@merl.com),<br />
Ahmed Hussen AbdelAziz (Ahmed.HussenAbdelAziz@rub.de),<br />
Dorothea Kolossa (dorothea.kolossa@rub.de)</div>Ramonhttps://www.hlt.inesc-id.pt/wiki/index.php?title=Robust_Speech_Processing_using_Observation_Uncertainty_and_Uncertainty_Propagation&diff=7459Robust Speech Processing using Observation Uncertainty and Uncertainty Propagation2015-01-07T17:04:08Z<p>Ramon: </p>
<hr />
<div>== Call for Papers for a Special Session at Interspeech 2015 *Deadline 20 March* ==<br />
<br />
We are pleased to announce a special session at [http://interspeech2015.org/ Interspeech 2015] in Dresden. The session will focusi on uncertainty for robust speech processing across all sub-fields. This would cover <br />
<br />
* Methods exploiting observation and parameter uncertainty as well as uncertainty propagation for robust speech processing e.g. for inference, learning or adaptation<br />
<br />
* Especially welcome are papers on the topic deep learning and uncertainty. We are confident this is the right moment to tackle this issue and already expect works on the topic<br />
<br />
* All sub-fields are wellcome. Not only ASR and speaker recognition communities but also exploratory works on e.g. paralinguistics or transcription level uncertainties (lattices).<br />
<br />
In order to motivate exploratory ideas, and facilitate research we will provide the following tools to those interested in participating:<br />
<br />
* Feature extractors for different types of uncertainties (Wiener posterior, sparsity assumption, Supergaussian) compatible with HTK and Kaldi in Matlab<br />
<br />
* Patches for uncertainty decoding and modified imputation in HTK, thus allowing to use uncertainties in the existing recipes.<br />
<br />
* Patches to read Matlab produced DNN posteriors with Kaldi, thus allowing to implement uncertainty techniques based on DNNs in Matlab. <br />
<br />
The use of these tools is *not* obligatory. They are just provided for those interested.<br />
<br />
Ramón Fernandez Astudillo (ramon.astudillo@inesc-id.pt),<br />
Shinji Watanabe (watanabe@merl.com),<br />
Ahmed Hussen AbdelAziz (Ahmed.HussenAbdelAziz@rub.de),<br />
Dorothea Kolossa (dorothea.kolossa@rub.de)</div>Ramonhttps://www.hlt.inesc-id.pt/wiki/index.php?title=Robust_Speech_Processing_using_Observation_Uncertainty_and_Uncertainty_Propagation&diff=7458Robust Speech Processing using Observation Uncertainty and Uncertainty Propagation2015-01-07T11:05:38Z<p>Ramon: </p>
<hr />
<div>== Call for Papers for a Special Session at Interspeech 2015 *Deadline 20 March* ==<br />
<br />
We are pleased to announce a special session at Interspeech 2015 to take place on the 6-10 of September in Dresden, focusing on uncertainty for robust speech processing across all sub-fields. This would cover <br />
<br />
* Methods exploiting observation and parameter uncertainty as well as uncertainty propagation for robust speech processing e.g. for inference, learning or adaptation<br />
<br />
* Especially welcome are papers on the topic deep learning and uncertainty. We are confident this is the right moment to tackle this issue and already expect works on the topic<br />
<br />
* All sub-fields are wellcome. Not only ASR and speaker recognition communities but also exploratory works on e.g. paralinguistics or transcription level uncertainties (lattices).<br />
<br />
In order to motivate exploratory ideas, and facilitate research we will provide the following tools to those interested in participating:<br />
<br />
* Feature extractors for different types of uncertainties (Wiener posterior, sparsity assumption, Supergaussian) compatible with HTK and Kaldi in Matlab<br />
<br />
* Patches for uncertainty decoding and modified imputation in HTK, thus allowing to use uncertainties in the existing recipes.<br />
<br />
* Patches to read Matlab produced DNN posteriors with Kaldi, thus allowing to implement uncertainty techniques based on DNNs in Matlab. <br />
<br />
The use of these tools is *not* obligatory. They are just provided for those interested.<br />
<br />
Ramón Fernandez Astudillo (ramon.astudillo@inesc-id.pt),<br />
Shinji Watanabe (watanabe@merl.com),<br />
Ahmed Hussen AbdelAziz (Ahmed.HussenAbdelAziz@rub.de),<br />
Dorothea Kolossa (dorothea.kolossa@rub.de)</div>Ramonhttps://www.hlt.inesc-id.pt/wiki/index.php?title=Robust_Speech_Processing_using_Observation_Uncertainty_and_Uncertainty_Propagation&diff=7457Robust Speech Processing using Observation Uncertainty and Uncertainty Propagation2015-01-07T11:00:08Z<p>Ramon: Created page with "== Special Session at Interspeech 2015, 6-10 September in Dresden, *Deadline 20 March* == We are pleased to announce a special session at Interspeech 2015 focusing on uncertain..."</p>
<hr />
<div>== Special Session at Interspeech 2015, 6-10 September in Dresden, *Deadline 20 March* ==<br />
<br />
We are pleased to announce a special session at Interspeech 2015 focusing on uncertainty for robust speech processing across all sub-fields. This would cover <br />
<br />
* Methods exploiting observation and parameter uncertainty as well as uncertainty propagation for robust speech processing e.g. for inference, learning or adaptation<br />
<br />
* Especially welcome are papers on the topic deep learning and uncertainty. We are confident this is the right moment to tackle this issue and already expect works on the topic<br />
<br />
* All sub-fields are wellcome. Not only ASR and speaker recognition communities but also exploratory works on e.g. paralinguistics or transcription level uncertainties (lattices).<br />
<br />
In order to motivate exploratory ideas, and facilitate research we will provide the following tools to those interested in participating:<br />
<br />
* Feature extractors for different types of uncertainties (Wiener posterior, sparsity assumption, Supergaussian) compatible with HTK and Kaldi in Matlab<br />
<br />
* Patches for uncertainty decoding and modified imputation in HTK, thus allowing to use uncertainties in the existing recipes.<br />
<br />
* Patches to read Matlab produced DNN posteriors with Kaldi, thus allowing to implement uncertainty techniques based on DNNs in Matlab. <br />
<br />
The use of these tools is *not* obligatory. They are just provided for those interested.<br />
<br />
Ramón Fernandez Astudillo (ramon.astudillo@inesc-id.pt),<br />
Shinji Watanabe (watanabe@merl.com),<br />
Ahmed Hussen AbdelAziz (Ahmed.HussenAbdelAziz@rub.de),<br />
Dorothea Kolossa (dorothea.kolossa@rub.de)</div>Ramonhttps://www.hlt.inesc-id.pt/wiki/index.php?title=Ramon_Fernandez_Astudillo&diff=7456Ramon Fernandez Astudillo2015-01-07T10:55:52Z<p>Ramon: </p>
<hr />
<div>{{infobox|name=Ramón Fernandez Astudillo<br />
|username=ramon<br />
|contact=ramon.astudillo<br />
|phone=+351-213-100-315<br />
|fax=+351-213-145-843<br />
}}<br />
<!-- YOUR SHORT BIO SHOULD BE HERE --><br />
I obtained the industrial engineering degree with specialization electronics in automatic regulation at the ''Escuela Politecnica Superior de Ingenieria de Gijon'' (Spain) in 2005, completing the last two years of this degree with an Erasmus scholarship at the ''Technische Universität Berlin''. In 2006 I worked as an intern at Peiker Acustic researching model-based speech enhancement. On this same year I was awarded with a ''La Caixa'' and the German Academic Exchange Service (DAAD) scholarship for research towards the Ph.D. degree. I obtained the title with distinction from the ''Technische Universität Berlin'' in 2010 in the fields of speech processing and robust automatic speech recognition. In my thesis "Integration of Short-Time Fourier Domain Speech Enhancement and Observation Uncertainty Techniques for Robust Automatic Speech Recognition" I formalized the concept of short-time Fourier uncertainty propagation and introduced various closed from solutions for it. I am currently a Post.-Doc. researcher here at INESC-ID/L2F researching both on robust speech recognition and robust natural language processing speech applications in a Bayesian setting. I am an ISCA member and reviewer of IEEE-TASLP/SPL as well as CSL. My interests span statistics, signal processing, machine learning and in general both human and machine perception of the environment.<br />
<br />
<inesc-id what='person' id='1289'></inesc-id><br />
<br />
<br />
== Special Session at Interspeech 2015==<br />
<br />
[[Robust Speech Processing using Observation Uncertainty and Uncertainty Propagation]]<br />
<br />
== Old Publications ==<br />
<br />
A complete list of my publications can be found [http://www.astudillo.com/ramon/research/ here], in particular, my doctoral thesis can be found [http://opus.kobv.de/tuberlin/volltexte/2010/2676/pdf/astudillo_ramon.pdf here]<br />
<br />
== Around the Web ==<br />
My LinkedIn profile can be found [http://www.linkedin.com/in/ramonastudillo here] (I only add people I meet in person).<br />
<br />
[[category:People]]<br />
[[category:Researchers]]</div>Ramonhttps://www.hlt.inesc-id.pt/wiki/index.php?title=Ramon_Fernandez_Astudillo&diff=7455Ramon Fernandez Astudillo2015-01-06T17:09:52Z<p>Ramon: </p>
<hr />
<div>{{infobox|name=Ramón Fernandez Astudillo<br />
|username=ramon<br />
|contact=ramon.astudillo<br />
|phone=+351-213-100-315<br />
|fax=+351-213-145-843<br />
}}<br />
<!-- YOUR SHORT BIO SHOULD BE HERE --><br />
I obtained the industrial engineering degree with specialization electronics in automatic regulation at the ''Escuela Politecnica Superior de Ingenieria de Gijon'' (Spain) in 2005, completing the last two years of this degree with an Erasmus scholarship at the ''Technische Universität Berlin''. In 2006 I worked as an intern at Peiker Acustic researching model-based speech enhancement. On this same year I was awarded with a ''La Caixa'' and the German Academic Exchange Service (DAAD) scholarship for research towards the Ph.D. degree. I obtained the title with distinction from the ''Technische Universität Berlin'' in 2010 in the fields of speech processing and robust automatic speech recognition. In my thesis "Integration of Short-Time Fourier Domain Speech Enhancement and Observation Uncertainty Techniques for Robust Automatic Speech Recognition" I formalized the concept of short-time Fourier uncertainty propagation and introduced various closed from solutions for it. I am currently a Post.-Doc. researcher here at INESC-ID/L2F researching both on robust speech recognition and robust natural language processing speech applications in a Bayesian setting. I am an ISCA member and reviewer of IEEE-TASLP/SPL as well as CSL. My interests span statistics, signal processing, machine learning and in general both human and machine perception of the environment.<br />
<br />
<inesc-id what='person' id='1289'></inesc-id><br />
<br />
<br />
== Special Session at Interspeech 2015==<br />
<br />
[[Special Session on Robust Speech Processing using Observation Uncertainty and Uncertainty Propagation]]<br />
<br />
== Old Publications ==<br />
<br />
A complete list of my publications can be found [http://www.astudillo.com/ramon/research/ here], in particular, my doctoral thesis can be found [http://opus.kobv.de/tuberlin/volltexte/2010/2676/pdf/astudillo_ramon.pdf here]<br />
<br />
== Around the Web ==<br />
My LinkedIn profile can be found [http://www.linkedin.com/in/ramonastudillo here] (I only add people I meet in person).<br />
<br />
[[category:People]]<br />
[[category:Researchers]]</div>Ramonhttps://www.hlt.inesc-id.pt/wiki/index.php?title=Ramon_Fernandez_Astudillo&diff=7454Ramon Fernandez Astudillo2015-01-06T17:08:53Z<p>Ramon: </p>
<hr />
<div>{{infobox|name=Ramón Fernandez Astudillo<br />
|username=ramon<br />
|contact=ramon.astudillo<br />
|phone=+351-213-100-315<br />
|fax=+351-213-145-843<br />
}}<br />
<!-- YOUR SHORT BIO SHOULD BE HERE --><br />
I obtained the industrial engineering degree with specialization electronics in automatic regulation at the ''Escuela Politecnica Superior de Ingenieria de Gijon'' (Spain) in 2005, completing the last two years of this degree with an Erasmus scholarship at the ''Technische Universität Berlin''. In 2006 I worked as an intern at Peiker Acustic researching model-based speech enhancement. On this same year I was awarded with a ''La Caixa'' and the German Academic Exchange Service (DAAD) scholarship for research towards the Ph.D. degree. I obtained the title with distinction from the ''Technische Universität Berlin'' in 2010 in the fields of speech processing and robust automatic speech recognition. In my thesis "Integration of Short-Time Fourier Domain Speech Enhancement and Observation Uncertainty Techniques for Robust Automatic Speech Recognition" I formalized the concept of short-time Fourier uncertainty propagation and introduced various closed from solutions for it. I am currently a Post.-Doc. researcher here at INESC-ID/L2F researching both on robust speech recognition and robust natural language processing speech applications in a Bayesian setting. I am an ISCA member and reviewer of IEEE-TASLP/SPL as well as CSL. My interests span statistics, signal processing, machine learning and in general both human and machine perception of the environment.<br />
<br />
<inesc-id what='person' id='1289'></inesc-id><br />
<br />
<br />
== Special Session at Interspeech 2015==<br />
<br />
[[Robust Speech Processing using Observation Uncertainty and Uncertainty Propagation]]<br />
<br />
== Old Publications ==<br />
<br />
A complete list of my publications can be found [http://www.astudillo.com/ramon/research/ here], in particular, my doctoral thesis can be found [http://opus.kobv.de/tuberlin/volltexte/2010/2676/pdf/astudillo_ramon.pdf here]<br />
<br />
== Around the Web ==<br />
My LinkedIn profile can be found [http://www.linkedin.com/in/ramonastudillo here] (I only add people I meet in person).<br />
<br />
[[category:People]]<br />
[[category:Researchers]]</div>Ramon