Revision as of 16:42, 3 February 2014

Sponsored by: FCT (PTDC/EIA-CCO/122542/2010)
Start: February 2012
End: January 2015

Team

Project Leader: Isabel Trancoso

• Alberto Abad
• António Serralheiro
• João Neto
• João Paulo Carvalho
• Luís Oliveira
• José Portêlo
• Carlos Ribeiro
• Rui Joaquim
• Gonçalo Carito

Consultant: Bhiksha Raj

Summary

The goal of this project is to develop privacy-preserving frameworks for processing voice data. Processing will be performed without having access to the voice, i.e., access to any form of the speech that can be analyzed to obtain information about the talker or what they spoke. Using a combination of tools from cryptography and secure-multiparty computation we will render voice processing algorithms secure (i.e. privacy-preserving), so that the privacy of all parties is preserved. We specifically propose to develop solutions for secure speaker verification and keyword spotting problems. We envision a mechanism whereby police/security agencies could obtain publicly- accepted forms of legal sanction to look for pre-specified voices or phrases. The privacy-preserving framework will ensure that they are only notified when these occur, but will have no access to the voice data itself, thus preserving citizens’ privacy. Although this proposal addresses only speech processing techniques, the impact of this type of technique is not restricted to this area. In fact, it may affect all types of multimedia documents.

Workplan

• T1 - Secure multi-party computation
• T2 - Secure speech parametrization
• T3 - Secure speaker verification
• T4 - Secure keyword spotting
• T5 - Proof of concept
• T6 - Evaluation
• T7 - Management and dissemination

Publications

Maria Joana Correia, Towards an Anti-spoofing System Based on Phase and Modulation Features, INESC-ID Technical Report RT/28/2013, December 2013

Alberto Abad and Ramon Astudillo and Isabel Trancoso, The L2F Spoken Web Search system for Mediaeval 2013, In Proc. Mediaeval Workshop, Barcelona, Spain, October 2013

José Portelo, Alberto Abad, Bhiksha Raj, Isabel Trancoso, Secure binary embeddings of front-end factor analysis for privacy preserving speaker verification, In INTERSPEECH 2013, Lyon, France, August 2013

José Portelo, Bhiksha Raj, Petros Boufounos, Isabel Trancoso, Alberto Abad, Speaker verification using secure binary embeddings, In EUSIPCO 2013, Marrakech, Morocco, September 2013

Rui Joaquim, Paulo Ferreira, Carlos Ribeiro, EVIV: an End-to-end Verifiable Internet Voting System, In Computer & Security, 32, pp. 170-191, Jan. 2013, Elsevier

Mauro Silva, João Leitão, Carlos Ribeiro, MobUser: Publish-Subscribe Communication for Mobile Nodes, Proc. 18th IEEE International Conference on Parallel and Distributed Systems, Dec. 2012

Manas Pathak, José Portelo, Bhiksha Raj, Isabel Trancoso, Privacy-preserving speaker authentication, In ISC 2012, Springer-Verlag , series LNCS 7483, pages 1-22, Passau, Germany, September 2012 (Keynote Talk by Bhiksha Raj)

José Portelo, Bhiksha Raj, Isabel Trancoso, Attacking a Privacy Preserving Music Matching Algorithm, In ICASSP 2012, IEEE, Kyoto, Japan, March 2012

José Portelo, Bhiksha Raj, Alberto Abad, Isabel Trancoso, On the Implementation of a Secure Musical Database Matching, In Proc. EUSIPCO 2011 - 19th European Signal Processing Conference, EURASIP, Barcelona, Spain, August 2011

Seminars

Bhiksha Raj, Unsupervised semantic structure discovery for audio, INESC-ID, October 31st, 2013

José Portelo, Privacy Preserving Speaker Verification using Secure Binary Embeddings, INESC-ID, October 4th, 2013

Bhiksha Raj, Privacy-Preserving Speech and Audio Processing, INESC-ID, September 28th 2012

Luís Brandão, 1-output 2-party secure function evaluation in a malicious setting via a cut and choose of garbled circuits, INESC-ID, October 2nd 2012

See also

Carnegie Mellon Voice Verification Technology Prevents Impersonators From Obtaining Voiceprints, Press Release, September 2012

David Talbot, Wiping away your Siri fingerprint, MIT Technology Review, June 2012