Domain-adversarial training for session independent EMG-based speech recognition

Michael Wand, Tanja Schultz, Jürgen Schmidhuber

Research output: Chapter in Book/Report/Conference proceedingConference contribution

14 Scopus citations

Abstract

We present our research on continuous speech recognition based on Surface Electromyography (EMG), where speech information is captured by electrodes attached to the speaker's face. This method allows speech processing without requiring that an acoustic signal is present; however, reattachment of the EMG electrodes causes subtle changes in the recorded signal, which degrades the recognition accuracy and thus poses a major challenge for practical application of the system. Based on the growing body of recent work in domain-adversarial training of neural networks, we present a system which adapts the neural network frontend of our recognizer to data from a new recording session, without requiring supervised enrollment.
Original languageEnglish (US)
Title of host publicationProceedings of the Annual Conference of the International Speech Communication Association, INTERSPEECH
PublisherInternational Speech Communication Association4 Rue des Fauvettes - Lous TourilsBaixas66390publication@isca-speech.org
Pages3167-3171
Number of pages5
DOIs
StatePublished - Jan 1 2018
Externally publishedYes

Fingerprint

Dive into the research topics of 'Domain-adversarial training for session independent EMG-based speech recognition'. Together they form a unique fingerprint.

Cite this