Recognition of Reverberant Speech using Full Cepstral Features and Spectral Missing Data.
Kalle J. PalomäkiGuy J. BrownJon P. BarkerPublished in: ICASSP (1) (2006)
Keyphrases
- missing data
- cepstral features
- feature extraction
- speech signal
- missing values
- matrix factorization
- incomplete data
- low rank
- structure from motion
- pattern recognition
- speech sounds
- speech recognition
- motion segmentation
- recognition algorithm
- data imputation
- automatic speech recognition
- missing information
- dirichlet process mixture models
- spectral features
- face recognition