Nonlinear enhancement of noisy speech, using continuous attractor dynamics formed in recurrent neural networks.
Louiza DehyadegarySeyyed Ali SeyyedsalehiIsar NejadgholiPublished in: Neurocomputing (2011)
Keyphrases
- recurrent neural networks
- recurrent networks
- nonlinear dynamic systems
- echo state networks
- dynamical systems
- neural model
- noisy speech
- feed forward
- neural network
- dynamic behavior
- hidden markov models
- cellular automata
- artificial neural networks
- reservoir computing
- speech signal
- noisy environments
- speech enhancement
- automatic speech recognition systems
- phase space
- image enhancement
- speech recognition
- back propagation
- image processing
- signal to noise ratio