Unsupervised learning of phonemes of whispered speech in a noisy environment based on convolutive non-negative matrix factorization.
Jian ZhouRuiyu LiangLi ZhaoLiang TaoCairong ZouPublished in: Inf. Sci. (2014)
Keyphrases
- negative matrix factorization
- noisy environments
- unsupervised learning
- speech signal
- speech recognition
- automatic speech recognition
- speech synthesis
- speaker identification
- sparse representation
- matrix factorization
- speech enhancement
- speaker verification
- endpoint detection
- background noise
- document clustering
- dimensionality reduction
- hidden markov models
- principal component analysis
- data clustering
- supervised learning
- vocal tract
- spectral clustering
- voice activity detection
- semi supervised
- non stationary
- expectation maximization
- language model
- pattern recognition
- machine learning
- neural network
- feature selection
- cluster analysis
- noise reduction
- text classification
- named entities
- document collections
- signal processing
- object recognition
- information retrieval