New analytical calculation and estimation of TDOA for underdetermined BSS in noisy environments.
Takuro MaruyamaShoko ArakiTomohiro NakataniShigeki MiyabeTakeshi YamadaShoji MakinoAtsushi NakamuraPublished in: APSIPA (2012)
Keyphrases
- noisy environments
- blind source separation
- noise reduction
- speech recognition
- speaker verification
- speech signal
- speech enhancement
- mixing matrix
- background noise
- digit recognition
- speaker identification
- independent component analysis
- non stationary
- probabilistic model
- multiresolution
- feature vectors
- computer vision