CASA-based speaker identification using cascaded GMM-CNN classifier in noisy and emotional talking conditions.
Ali Bou NassifIsmail ShahinShibani HamsaNawel NemmourKeikichi HirosePublished in: Appl. Soft Comput. (2021)
Keyphrases
- speaker identification
- noisy environments
- probabilistic neural network
- gaussian mixture model
- speaker recognition
- speech recognition
- speech signal
- feature extraction
- noise reduction
- speaker verification
- training data
- feature space
- broadcast news
- audio signals
- mel frequency cepstral coefficients
- automatic speech recognition
- mixture model
- feature selection
- feature set
- pattern recognition
- support vector
- neural network
- expectation maximization
- low level
- image retrieval
- training set