Predictive deconvolution and kurtosis maximization for speech dereverberation.
David T. FeeC. F. N. CowanStefan BilbaoIzzet OzcelikPublished in: EUSIPCO (2006)
Keyphrases
- speech recognition
- speech signal
- least squares
- standard deviation
- recognition engine
- audio visual
- image restoration
- speech synthesis
- objective function
- automatic speech recognition
- image deconvolution
- higher order statistics
- endpoint detection
- motion deblurring
- text to speech
- speaker recognition
- speaker identification
- spoken language
- inverse problems
- predictive modeling
- small size
- predictive model
- denoising
- broadcast news
- blind deconvolution
- emotion recognition
- linear prediction
- language acquisition
- multimodal interfaces
- distribution function
- vocal tract
- emotional state