An Investigation of Subband Wavenet Vocoder Covering Entire Audible Frequency Range with Limited Acoustic Features.
Takuma OkamotoKentaro TachibanaTomoki TodaYoshinori ShigaHisashi KawaiPublished in: ICASSP (2018)
Keyphrases
- subband
- acoustic features
- automatic speech recognition
- filter bank
- wavelet transform
- image compression
- multiresolution
- bit rate
- wavelet coefficients
- feature vectors
- speaker verification
- subband coding
- frequency domain
- discrete wavelet transform
- speech signal
- noisy environments
- image coder
- subband decomposition
- music information retrieval
- wavelet subbands
- visual features
- cross correlation
- hidden markov models
- image coding
- rate distortion
- vector quantization
- audio features
- perfect reconstruction
- video sequences
- multiscale
- image sequences
- image processing