MSMC-TTS: Multi-Stage Multi-Codebook VQ-VAE Based Neural TTS.
Haohan GuoFenglong XieXixin WuFrank K. SoongHelen MengPublished in: IEEE ACM Trans. Audio Speech Lang. Process. (2023)
Keyphrases
- multistage
- text to speech
- vector quantization
- vector quantizer
- image compression
- finite state vector quantization
- production system
- dynamic programming
- single stage
- lot sizing
- stochastic optimization
- vector quantized
- image coding
- stochastic programming
- network architecture
- neural network
- bag of words
- word processing
- image representation
- optimal policy
- attack detection
- codebook design
- codebook generation
- wavelet transform
- lot streaming
- np hard