Privacy-Preserving Speech Representation Learning using Vector Quantization

March 15, 2022 Β· Declared Dead Β· πŸ› XXXIVe JournΓ©es d'Γ‰tudes sur la Parole -- JEP 2022

πŸ‘» CAUSE OF DEATH: Ghosted
No code link whatsoever

"No code URL or promise found in abstract"

Evidence collected by the PWNC Scanner

Authors Pierre Champion, Denis Jouvet, Anthony Larcher arXiv ID 2203.09518 Category eess.AS: Audio & Speech Cross-listed cs.AI, cs.CL, cs.CR, cs.SD Citations 0 Venue XXXIVe JournΓ©es d'Γ‰tudes sur la Parole -- JEP 2022 Last Checked 3 months ago
Abstract
With the popularity of virtual assistants (e.g., Siri, Alexa), the use of speech recognition is now becoming more and more widespread.However, speech signals contain a lot of sensitive information, such as the speaker's identity, which raises privacy concerns.The presented experiments show that the representations extracted by the deep layers of speech recognition networks contain speaker information.This paper aims to produce an anonymous representation while preserving speech recognition performance.To this end, we propose to use vector quantization to constrain the representation space and induce the network to suppress the speaker identity.The choice of the quantization dictionary size allows to configure the trade-off between utility (speech recognition) and privacy (speaker identity concealment).
Community shame:
Not yet rated
Community Contributions

Found the code? Know the venue? Think something is wrong? Let us know!

πŸ“œ Similar Papers

In the same crypt β€” Audio & Speech

Died the same way β€” πŸ‘» Ghosted