Exploiting the potential of auditory preprocessing for robust speech recognition by locally recurrent neural networks.
Klaus KasperHerbert ReiningerDietrich WolfPublished in: ICASSP (1997)
Keyphrases
- speech recognition
- recurrent neural networks
- preprocessing
- noisy environments
- hidden markov models
- speech understanding
- feed forward
- pattern recognition
- neural network
- speech recognition technology
- speech synthesis
- speech processing
- speaker identification
- reservoir computing
- echo state networks
- automatic speech recognition
- speech signal
- language model
- speech recognizer
- recurrent networks
- signal processing
- speech recognition systems
- nonlinear dynamic systems
- speaker dependent
- artificial neural networks
- word recognition
- speaker recognition
- speaker independent
- feature extraction
- isolated word
- speech recognizers
- cepstral coefficients
- back propagation
- background noise