EMGSE: Acoustic/EMG Fusion for Multimodal Speech Enhancement.
Kuan-Chen WangKai-Chun LiuHsin-Min WangYu TsaoPublished in: CoRR (2022)
Keyphrases
- speech enhancement
- noisy speech
- noise reduction
- noisy environments
- signal to noise ratio
- speech signal
- single channel
- hidden markov models
- linear prediction
- background noise
- multi channel
- acoustic features
- multi modal
- edge preserving
- speech recognition
- smoothing algorithm
- vocal tract
- non stationary
- wiener filter
- speaker verification