Speech emotion recognition using multichannel parallel convolutional recurrent neural networks based on gammatone auditory filterbank.
Zhichao PengZhi ZhuMasashi UnokiJianwu DangMasato AkagiPublished in: APSIPA (2017)
Keyphrases
- recurrent neural networks
- filter bank
- signal processing
- linear prediction
- image coding
- subband
- emotion recognition
- feed forward
- neural network
- echo state networks
- recurrent networks
- multiscale
- spectral analysis
- emotional state
- text to speech synthesis
- speech recognition
- artificial neural networks
- multiresolution
- feedforward neural networks
- perfect reconstruction
- cascade correlation
- subband coding
- neural model
- reservoir computing
- frequency domain
- wavelet basis
- audio visual
- facial expressions
- visual information
- speech signal
- control system
- pattern recognition
- chaotic time series
- wavelet bases
- computationally efficient
- wavelet coefficients