Model based binaural enhancement of voiced and unvoiced speech.
Mathew Shaji KavalekalamMads Græsbøll ChristensenJesper Bünsow BoldtPublished in: ICASSP (2017)
Keyphrases
- speech signal
- sound source
- speech recognition
- speech quality
- automatic speech recognition
- vocal tract
- speaker identification
- non stationary
- automatic speech recognition systems
- background noise
- hidden markov models
- speech processing
- spectral analysis
- linear prediction
- noisy environments
- image enhancement
- fundamental frequency
- speaker recognition
- emotion recognition
- transfer function
- data driven
- fully unsupervised
- data sets
- contrast enhancement
- natural language