EMOQ-TTS: Emotion Intensity Quantization for Fine-Grained Controllable Emotional Text-to-Speech.
Chae-Bin ImSang-Hoon LeeSeung-Bin KimSeong-Whan LeePublished in: ICASSP (2022)
Keyphrases
- text to speech
- fine grained
- text to speech synthesis
- emotional state
- emotion recognition
- coarse grained
- speech synthesis
- affective computing
- prosodic features
- tightly coupled
- human decision making
- word processing
- access control
- human behaviour
- facial expressions
- programming tool
- affective states
- virtual characters
- massively parallel
- sentiment analysis
- databases
- co occurrence
- hidden markov models
- affect detection
- data lineage
- database systems