Ramon Fernandez Astudillo


Ramón Fernandez Astudillo

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 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 (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 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 (RoSP-SIG) and organizer of the LxMLS summer schools 2015, 2016.

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.

You can find my GitHub repo here


Book Chapters


  • J. M. Lemos, Ramon Fernandez Astudillo, Angela Costa, André Silva, Hugo Silva, João Graça, Bhiksha Raj, Crowdsourced Video Subtitling with Adaptation Based on User-Corrected Lattices, September 2016

International Journals


  • Andre Martins, Marcin Junczys-Dowmunt, Fábio Natanael Kepler, Ramon Fernandez Astudillo, Chris Hokamp, Roman Grundkiewicz, Pushing the Limits of Translation Quality Estimation, Transactions of the Association for Computational Linguistics (TACL), January 2017


  • Rahim Saedi, Ramon Fernandez Astudillo, Dorothea Kolossa, Uncertain LDA: Including observation uncertainties in discriminative transforms, IEEE Transactions on Pattern Analysis and Machine Intelligence, IEEE, August 2015


  • Dorothea Kolossa, Steffen Zeiler, Rahim Saedi, Ramon Fernandez Astudillo, Noise-adaptive LDA: A new approach for speech recognition under observation uncertainty, IEEE SIGNAL PROCESSING LETTERS, IEEE, vol. 20, n. 11, pages 1018 - 1021, doi: 10.1109/LSP.2013.2278556, August 2013
  • Ramon Fernandez Astudillo, Reinhold Orglmeister, Computing MMSE Estimates and Residual Uncertainty directly in the Feature Domain of ASR using STFT Domain Speech Distortion Models, IEEE TRANSACTIONS ON AUDIO, SPEECH, AND LANGUAGE PROCESSING, IEEE, vol. 21, n. 5, pages 1023-1034, May 2013
  • Robert M. Nickel, Ramon Fernandez Astudillo, Dorothea Kolossa, Rainer Martin, Corpus-Based Speech Enhancement With Uncertainty Modeling and Cepstral Smoothing, IEEE TRANSACTIONS ON AUDIO, SPEECH, AND LANGUAGE PROCESSING, IEEE, vol. 21, n. 5, pages 983-997, May 2013
  • Ramon Fernandez Astudillo, Dorothea Kolossa, Alberto Abad, Steffen Zeiler, Rahim Saeidi, Pejman Mowlaee, João Paulo da Silva Neto, Rainer Martin, Integration of beamforming and uncertainty-of-observation techniques for robust ASR in multi-source environments, Computer Speech and Language, vol. 27, n. 3, pages 837-850, May 2013
  • Ramon Fernandez Astudillo, An Extension of STFT Uncertainty Propagation for GMM-Based Super-Gaussian a Priori Models, IEEE SIGNAL PROCESSING LETTERS, IEEE, vol. 20, n. 12, pages 1163-1166, December 2013

International Conferences


  • Lucia Specia, Frederic Blain, Varvara Logacheva, Ramon Fernandez Astudillo, Andre Martins, Findings of the wmt 2018 shared task on quality estimation, In Proceedings of the Third Conference on Machine Translation: Shared Task Papers, October 2018


  • Fábio Natanael Kepler, Ramon Fernandez Astudillo, Alberto Abad, Fusion of Simple Models for Native Language Identification, In 12th Workshop on Innovative Use of NLP for Building Educational Applications, Association for Computational Linguistics, pages 423-429, doi: 10.18653/v1/W17-5048, September 2017
  • Miguel Varela Ramos, Alan Black, Ramon Fernandez Astudillo, Isabel Trancoso, Nuno Fonseca, Segment Level Voice Conversion with Recurrent Neural Networks, In INTERSPEECH 2017, ISCA, August 2017
  • Christian Huemmer, Ramon Fernandez Astudillo, Walter Kellerman, An improved uncertainty decoding scheme with weighted samples for multi-channel DNN-HMM hybrid systems, In Hands-free Speech Communications and Microphone Arrays (HSCMA), March 2017


  • André T. Martins, Ramon Fernandez Astudillo, Chris Hokamp, Fábio Natanael Kepler, Unbabel’s Participation in the WMT16 Word-Level Translation Quality Estimation Shared Task, August 2016
  • André T. Martins, Ramon Fernandez Astudillo, From Softmax to Sparsemax: A Sparse Model of Attention and Multi-Label Classification, July 2016
  • Sílvio Moreira, Ramon Fernandez Astudillo, Wang Ling, Mário J. Silva, Isabel Trancoso, INESC-ID at SemEval-2016 Task 4-A: Reducing the Problem of Out-of-Embedding Words, In International Workshop on Semantic Evaluation (SemEval-2016), pages 238--242, July 2016


  • Christian Huemmer, Roland Maas, Andreas Schwarz, Ramon Fernandez Astudillo, Walter Kellermann, Uncertainty Decoding for DNN-HMM Hybrid Systems Based on Numerical Sampling, In Proceedings Interspeech 2015, ISCA, September 2015


  • Alberto Abad, José Matos, Hugo Meinedo, Ramon Fernandez Astudillo, Isabel Trancoso, The L2F system for the EVALITA-2014 speech activity detection challenge in domestic environments, In International Workshop EVALITA 2014, Pisa, Italy, December 2014
  • José Matos, Alberto Abad, Ramon Fernandez Astudillo, Isabel Trancoso, Recognition of distant voice commands for home applications in Portuguese, In Iberspeech 2014, Las Palmas de Gran Canaria, Spain, November 2014
  • Marco Matassoni, Ramon Fernandez Astudillo, Athanasios Katsamanis, Mirco Ravanelli, The DIRHA-GRID corpus: baseline and tools for multi-room distant speech recognition using distributed microphones, In Interspeech 2014, pages 1613-1617, September 2014


  • Ramon Fernandez Astudillo, A propagation approach to modelling the joint distributions of clean and corrupted speech in the Mel-cepstral domain, In IEEE Automatic Speech Recognition and Understanding Workshop (ASRU), pages 180-185, December 2013
  • Francesco Nesta, Marco Matassoni, Ramon Fernandez Astudillo, A FLEXIBLE SPATIAL BLIND SOURCE EXTRACTION FRAMEWORK FOR ROBUST SPEECH RECOGNITION IN NOISY ENVIRONMENTS, In 2nd International Workshop on Machine Listening in Multisource Environments (CHiME), pages 33-38, June 2013
  • Ramon Fernandez Astudillo, Timo Gerkmann, On the Relation Between Speech Corruption Models in the Spectral and the Cepstral Domain, In 38th International Conference on Acoustics, Speech, and Signal Processing (ICASSP), IEEE, pages 7044-7048, May 2013


  • R. M. Nickel, Ramon Fernandez Astudillo, D. Kolossa, S. Zeiler, R. Martin, Inventory-Style Speech Enhancement with Uncertainty-of-Observation Techniques, In ICASSP, pages 4645--4648, May 2012


  • Ramon Fernandez Astudillo, João Paulo da Silva Neto, Propagation of Uncertainty through Multilayer Perceptrons for Robust Automatic Speech Recognition, In Interspeech, pages 461-464, October 2011

National Journals


National Conferences


Other Publications


  • Ramon Fernandez Astudillo, Emmanuel Vincent, Li Deng, Uncertainty Handling for Environment-Robust Speech Recognition, Tutorial at Interspeech 2012 , September 2012

Special Session at Interspeech 2015

Robust Speech Processing using Observation Uncertainty and Uncertainty Propagation

Old Publications

A complete list of my publications can be found here, in particular, my doctoral thesis can be found here

Around the Web

My LinkedIn profile can be found here (I only add people I meet in person).