Reconstruction-based speech enhancement from robust acoustic features.
Philip HardingBen MilnerPublished in: Speech Commun. (2015)
Keyphrases
- noisy environments
- speech enhancement
- speech signal
- acoustic features
- speaker verification
- noise reduction
- super resolution
- speech recognition
- signal to noise ratio
- dynamic environments
- high resolution
- automatic speech recognition
- speaker recognition
- audio visual
- speaker identification
- vocal tract
- wiener filter
- audio features
- spectral analysis
- neural network
- non stationary
- multi modal
- pattern recognition
- bayesian networks
- machine learning