[Preview] DualVoice: A Speech Interaction Method Using Whisper Voice as Commands
[Preview] DualVoice: A Speech Interaction Method Using Whisper Voice as Commands
Jun Rekimoto
CHI'22: ACM Conference on Human Factors in Computing Systems
Session: Late Breaking Work (LBW) Virtual; Late Breaking Work (LBW)
Abstract
Applications based on speech recognition have become widely used, and speech input is increasingly being utilized to create documents. However, since there is no easy way to distinguish commands and text input in speech, it is still difficult to correct misrecognition by speech, which makes it necessary to re-edit documents by manual input. It is also difficult to input symbols and commands because these may be misrecognized as text letters. We propose a speech interaction method called DualVoice, in which commands are input in a whispered voice and letters in a normal voice. The proposed method does not require any special hardware other than a regular microphone, thus enabling a complete hands-free interaction. It can be used in a wide range of situations where speech recognition is already available. We designed two neural networks, one for discriminating normal speech from whispered speech, and the second for recognizing whisper speech.
WEB:: http://programs.sigchi.org/chi/2022/p...
Presentation Video:: • DualVoice: A Speech Interaction Metho...
DOI:: https://doi.org/10.1145/3491101.3519700
Video previews for CHI 2022 Late-Breaking Works
