Model-based independent component analysis for robust multi-microphone automatic speech recognition.
Laurent CouvreurChristophe RisPublished in: INTERSPEECH (2002)
Keyphrases
- automatic speech recognition
- independent component analysis
- speech recognition
- noisy environments
- independent components
- speech signal
- principal component analysis
- blind source separation
- factor analysis
- signal processing
- word error rate
- broadcast news
- hidden markov models
- spontaneous speech
- conversational speech
- dimensionality reduction
- speech corpus
- source separation
- face recognition
- high order statistics
- blind signal separation
- independent subspace analysis
- multi modal
- information retrieval systems
- probabilistic model
- single channel