My Lips Are Concealed: Audio-Visual Speech Enhancement Through Obstructions.
Triantafyllos AfourasJoon Son ChungAndrew ZissermanPublished in: INTERSPEECH (2019)
Keyphrases
- audio visual
- speech enhancement
- noisy environments
- noise reduction
- multi modal
- signal to noise ratio
- single channel
- speech signal
- linear prediction
- visual speech
- visual information
- visual data
- smoothing algorithm
- sound source
- speaker verification
- multimedia
- vocal tract
- wiener filter
- multi stream
- speech recognition
- speaker identification
- audio features
- information retrieval
- multi channel
- image classification