Sound source separation and automatic speech recognition for moving sources.
Kazuhiro NakadaiHirofumi NakajimaGökhan InceYuji HasegawaPublished in: IROS (2010)
Keyphrases
- source separation
- automatic speech recognition
- blind source separation
- speech signal
- acoustic features
- audio features
- speech recognition
- sound source
- single channel
- independent component analysis
- hidden markov models
- conversational speech
- denoising
- broadcast news
- temporal structure
- non stationary
- source signals
- low level
- audio visual
- speaker identification
- moving objects
- music information retrieval
- multi channel
- simple linear
- language model
- gaussian process
- machine learning
- image denoising
- model selection
- principal component analysis
- information extraction
- keywords
- face recognition
- image processing
- feature selection