News
CRISP paper on private speech recognition
Successful submissions from CRISP researchers
Several research groups in CRISP are concerned with private speech processing and recognition. Among them are the research group ENCRYPTO at TU Darmstadt headed by Prof. Thomas Schneider and the team led by Prof. Christoph Busch from Darmstadt University of Applied Sciences. The researchers from the TU Darmstadt have been able to place four papers at top conferences and journals, two of them were created in cooperation with Darmstadt University of Applied Sciences.
In these works, the researchers investigated different (existing and new) cryptographic approaches that can be used to recognize users based on their speech without violating their privacy or leaking their speech data. This can be useful for, e.g., banks that want to authenticate a client on the telephone using their speech while still protecting this sensitive biometric data. Additionally, the researchers of the ENCRYPTO group uncovered a significant flaw in a cryptographic protocol that has been, among users, used for prior work on private speaker recognition.
Accepted papers are:
Privacy-preserving PLDA speaker verification using outsourced secure computation
Authtors: Amos Treiber, Andreas Nautsch, Jascha Kolberg, Thomas Schneider, Christoph Busch
In: Speech Communication, 2019
PDF
A comment on privacy-preserving scalar product protocols as proposed in “SPOC”
Authors: Thomas Schneider, Amos Treiber
In: IEEE Transactions on Parallel and Distributed Systems (TPDS), 2019
PDF
Preserving privacy in speaker and speech characterisation
Authors: Andreas Nautsch, Abelino Jiménez, Amos Treiber, Jascha Kolberg, Catherine Jasserand, Els Kindt, Héctor Delgado, Massimiliano Todisco, Mohamed Amine Hmani, Aymen Mtibaa, Mohammed Ahmed Abdelraheem, Alberto Abad, Francisco Teixeira, Driss Matrouf, Marta Gomez-Barrero, Dijana Petrovska-Delacrétaz, Gérard Chollet, Nicholas Evans, Thomas Schneider, Jean-François Bonastre, Bhiksha Raj, Isabel Trancoso, Christoph Busch
In: Computer Speech and Language (CSL), 2019(58):441–480
PDF
Privacy-preserving speaker recognition with cohort score normalisation
Authors: Andreas Nautsch, Jose Patino, Amos Treiber, Themos Stafylakis, Petr Mizera, Massimiliano Todisco, Thomas Schneider, Nicholas Evans
In: 20. Annual Conference of the International Speech Communication Association (INTERSPEECH'19), International Speech Communication Association (ISCA)
PDF