A Spectral Energy Distance for Parallel Speech Synthesis.
Alexey A. GritsenkoTim SalimansRianne van den BergJasper SnoekNal KalchbrennerPublished in: NeurIPS (2020)
Keyphrases
- speech synthesis
- speech recognition
- text to speech
- vocal tract
- prosodic features
- distance function
- parallel processing
- energy consumption
- hidden markov models
- distance measure
- energy minimization
- speech corpus
- parallel programming
- language model
- image coding
- parallel implementation
- massively parallel
- distributed memory
- linear prediction
- graph cuts
- wavelet transform
- pattern recognition
- residual energy
- information retrieval
- neural network