Context dependent vector quantization for continuous speech recognition.
Lalit R. BahlPeter V. de SouzaP. S. GopalakrishnanMichael A. PichenyPublished in: ICASSP (2) (1993)
Keyphrases
- vector quantization
- context dependent
- continuous speech recognition
- image compression
- vector quantizer
- low level
- codebook design
- distortion measure
- fractal image compression
- input vector
- natural language
- speaker recognition
- bit allocation
- high level
- reduced complexity
- entropy constrained
- markov chain
- wavelet transform
- state space
- computational complexity
- finite state vector quantization