Discriminative Learning of Filterbank Layer within Deep Neural Network Based Speech Recognition for Speaker Adaptation.
Hiroshi SekiKazumasa YamamotoTomoyosi AkibaSeiichi NakagawaPublished in: IEICE Trans. Inf. Syst. (2019)
Keyphrases
- speech recognition
- speaker adaptation
- filter bank
- subband
- speaker independent
- speech recognizer
- speaker dependent
- automatic speech recognition
- hidden markov models
- multiscale
- multiresolution
- generative model
- speech synthesis
- language model
- neural network
- signal processing
- spectral analysis
- frequency domain
- maximum likelihood
- noisy environments
- speech signal
- pattern recognition
- wavelet transform
- probabilistic model
- speaker identification
- speech recognition systems
- bit rate
- mixture model
- multi modal
- video sequences
- image compression