MCNET: Fuse Multiple Cues for Multichannel Speech Enhancement.
Yujie YangChangsheng QuanXiaofei LiPublished in: ICASSP (2023)
Keyphrases
- speech enhancement
- multiple cues
- single channel
- linear prediction
- multi channel
- visual tracking
- independent component analysis
- prior information
- frequency domain
- noise reduction
- visual cues
- wiener filter
- particle filter
- noisy environments
- signal to noise ratio
- image descriptors
- object tracking
- video data
- visual information
- eye movements
- sound source
- lossless compression
- speech signal
- video frames