Perceive and Predict: Self-Supervised Speech Representation Based Loss Functions for Speech Enhancement.
George CloseWilliam RavenscroftThomas HainStefan GoetzePublished in: ICASSP (2023)
Keyphrases
- loss function
- speech enhancement
- speech signal
- noisy environments
- noise reduction
- signal to noise ratio
- vocal tract
- single channel
- pairwise
- loss minimization
- support vector
- linear prediction
- learning to rank
- convex loss functions
- speech recognition
- speech synthesis
- background noise
- higher order
- smoothing algorithm
- multi channel
- noisy speech
- maximum likelihood
- reproducing kernel hilbert space
- wiener filter
- markov random field
- risk minimization
- non stationary
- spatial domain
- maximum entropy