Spectral-Subtraction Based Features for Speaker Identification.
Mahesh ChandraPratibha NandiAparajita kumariShipra MishraPublished in: FICTA (2) (2014)
Keyphrases
- speaker identification
- speech recognition
- speech signal
- low level
- mel frequency cepstral coefficients
- feature set
- feature vectors
- prior knowledge
- computer vision
- gaussian mixture model
- feature extraction
- speaker recognition
- noisy images
- multi modal
- noisy environments
- classification accuracy
- video sequences
- training data