Two speaker speech separation by LP residual weighting and harmonics enhancement.
P. KrishnamoorthyS. R. Mahadeva PrasannaPublished in: Int. J. Speech Technol. (2010)
Keyphrases
- speech recognition
- speaker recognition
- automatic speech recognition
- audio visual
- speaker identification
- speaker verification
- speech signal
- fundamental frequency
- speaker dependent
- prosodic features
- linear programming
- speaker diarization
- vocal tract
- automatic speech recognition systems
- speech synthesis
- image enhancement
- hidden markov models
- gaussian mixture model
- linear program
- image processing
- speech sounds
- multi modal
- sound source
- frequency domain
- automatic transcription
- audio stream
- text to speech
- speaker adaptation
- synthesized speech
- speech recognizer
- noisy environments
- emotion recognition
- speaker independent
- spontaneous speech
- phoneme recognition
- pattern recognition
- audio features
- acoustic models
- visual speech
- broadcast news
- weighting scheme
- maximum likelihood
- language model
- similarity measure
- speech recognition systems
- acoustic features
- radon transform
- visual information
- basis functions
- high frequency
- np hard
- objective function