Audio hash function based on non-negative matrix factorisation of mel-frequency cepstral coefficients.
Ning ChenHe-D. XiaoWanggen WanPublished in: IET Inf. Secur. (2011)
Keyphrases
- hash functions
- mel frequency cepstral coefficients
- speaker identification
- speaker recognition
- gaussian mixture model
- acoustic features
- linear predictive
- audio features
- speech signal
- feature vectors
- feature set
- speech recognition
- speaker diarization
- spectral features
- principal component analysis
- extracting features
- similarity search
- speaker verification
- feature extraction
- probabilistic neural network
- data distribution
- audio visual
- cepstral coefficients
- feature selection
- broadcast news
- music information retrieval
- noisy environments
- automatic speech recognition
- low level
- vector quantization
- mixture model
- visual features
- multiresolution
- hidden markov models
- video sequences
- face recognition
- multi modal
- classification accuracy