Prosody-controllable spontaneous TTS with neural HMMs.
Harm LamerisShivam MehtaGustav Eje HenterJoakim GustafsonÉva SzékelyPublished in: CoRR (2022)
Keyphrases
- text to speech
- hidden markov models
- speech synthesis
- multi stream
- neural network
- network architecture
- prosodic features
- word processing
- speech recognition
- associative memory
- neural fuzzy
- pattern recognition
- artificial neural
- nonlinear predictive control
- neural computation
- biologically plausible
- learning rules
- bio inspired
- anomaly detection
- neural model
- recurrent networks
- feature extraction
- image segmentation
- genetic algorithm
- neural information processing
- chinese named entity recognition
- real time