Linguistics Colloquium: Yossi Keshet

Yossi Keshet, BIU

Title: Artificial Neural Models for Speech Processing: From Speech Science to Security

Abstract: 

Speech Science refers to the study of production, transmission, and perception of speech. It relates to areas of research such as linguistics, cognition, and medicine. In the first part of the talk, I will review a series of works that are aimed at extremely accurate (human-level) measurements of phonetic parameters, and I will describe their contributions to the speech science community. 

The second part of my talk will be devoted to my recent work on the security, and specifically on Steganography. Steganography ("steganos" - concealed or covered plus "graphein" - writing) is the science of concealing messages inside other messages. It is generally used to convey concealed "secret" messages to recipients who are aware of its presence while keeping even their existence hidden from other unaware parties who only see the "public" or "carrier" message. I will propose the use of deep neural networks as learnable steganographic functions, that learn to exploit redundancies in audio data to conceal messages optimally. I will demonstrate quantitatively that the proposed method can both effectively hide secret messages into a carrier and recover them from the carrier. Qualitative experiments suggest that modifications to the carrier are unnoticeable by human listeners and that the decoded messages are highly intelligible as well.

Place: Building 507 room 106

Subscribe to our Telegram channel to get notified about upcoming talks and events