Audio-Visual Speech Enhancement Using Conditional Variational Auto-Encoders.
Mostafa SadeghiSimon LeglaiveXavier Alameda-PinedaLaurent GirinRadu HoraudPublished in: IEEE ACM Trans. Audio Speech Lang. Process. (2020)
Keyphrases
- audio visual
- speech enhancement
- noisy environments
- noise reduction
- multi modal
- single channel
- signal to noise ratio
- speech signal
- linear prediction
- visual information
- sound source
- vocal tract
- visual data
- image segmentation
- multimedia
- smoothing algorithm
- multi stream
- wiener filter
- edge detection
- background noise
- speech recognition
- optical flow
- multi channel
- multiscale
- information retrieval
- image coding
- computer vision
- independent component analysis
- image processing
- image restoration
- hidden markov models