Speaker-Dependent Bottleneck Features for Egyptian Arabic Speech Recognition.
Aleksei RomanenkoValentin MendelevPublished in: SPECOM (2016)
Keyphrases
- speech recognition
- speaker dependent
- speaker identification
- speaker independent
- speech recognition systems
- hidden markov models
- speech recognizer
- speech signal
- automatic speech recognition
- language identification
- language model
- mel frequency cepstral coefficients
- noisy environments
- speech synthesis
- pattern recognition
- feature extraction
- feature set
- low level
- feature selection
- broadcast news
- feature vectors
- probabilistic neural network
- feature space
- speaker adaptation
- image processing