Estimating competing speaker count for blind speech source separation.
Valentin AndreiHoria CucuAndi BuzoCorneliu BurileanuPublished in: SpeD (2015)
Keyphrases
- source separation
- audio visual
- audio features
- speech recognition
- automatic speech recognition
- speaker identification
- blind source separation
- speech signal
- independent component analysis
- multi modal
- speaker diarization
- denoising
- visual information
- temporal structure
- single channel
- image retrieval
- multi channel
- visual data
- signal processing
- language model
- feature set
- noisy environments
- simple linear
- sound source
- hidden markov models
- multimedia