Designing a unified speech/audio codec by adopting a single channel harmonic source separation module.
Sang-Wook ShinChang-Heon LeeHyen-O OhHong-Goo KangPublished in: ICASSP (2008)
Keyphrases
- source separation
- single channel
- audio features
- audio visual
- multi channel
- sound source
- low level
- signal processing
- multi modal
- music information retrieval
- visual features
- independent component analysis
- text data
- audio signal
- frequency domain
- multimedia
- feature set
- speech recognition
- wiener filter
- prior information
- wavelet decomposition
- speech signal
- blind source separation
- motion estimation
- visual data
- visual information
- image compression
- image classification
- multiscale