Fusing MFCC and LPC Features Using 1D Triplet CNN for Speaker Recognition in Severely Degraded Audio Signals.
Anurag ChowdhuryArun RossPublished in: IEEE Trans. Inf. Forensics Secur. (2020)
Keyphrases
- speaker recognition
- speaker identification
- audio signals
- mel frequency cepstral coefficients
- speech signal
- speech recognition
- centroid neural network
- audio signal
- feature extraction
- gaussian mixture model
- probabilistic neural network
- feature vectors
- speaker verification
- feature set
- vector quantization
- extracted features
- spectral features
- broadcast news
- hidden markov models
- image processing
- non stationary
- visual features
- audio features
- language model
- feature space