A Conditional Generative Model for Speech Enhancement.
Zengxi LiLi-Rong DaiYan SongIan McLoughlinPublished in: Circuits Syst. Signal Process. (2018)
Keyphrases
- generative model
- speech enhancement
- noisy environments
- noise reduction
- single channel
- signal to noise ratio
- probabilistic model
- speech signal
- linear prediction
- vocal tract
- bayesian framework
- smoothing algorithm
- em algorithm
- speech recognition
- background noise
- additive noise
- wiener filter
- prior knowledge
- semi supervised
- multi channel
- conditional random fields
- expectation maximization
- image coding
- image segmentation
- automatic speech recognition
- topic models
- high resolution
- language model