SkipConvNet: Skip Convolutional Neural Network for Speech Dereverberation Using Optimally Smoothed Spectral Mapping.
Vinay KothapallyWei XiaShahram GhorbaniJohn H. L. HansenWei XueJing HuangPublished in: INTERSPEECH (2020)
Keyphrases
- convolutional neural network
- spectral features
- face detection
- speech recognition
- spectral analysis
- linear predictive coding
- linear prediction
- speech signal
- speech synthesis
- speaker recognition
- hyperspectral
- spoken language
- automatic speech recognition
- text to speech
- hidden markov models
- neural network
- cepstral coefficients
- automatic speech recognition systems
- endpoint detection
- hyperspectral imagery
- broadcast news
- hyperspectral images
- pattern recognition
- video sequences
- information retrieval