MsEmoTTS: Multi-Scale Emotion Transfer, Prediction, and Control for Emotional Speech Synthesis.
Yi LeiShan YangXinsheng WangLei XiePublished in: IEEE ACM Trans. Audio Speech Lang. Process. (2022)
Keyphrases
- speech synthesis
- multiscale
- speech recognition
- emotional state
- text to speech
- control system
- prediction accuracy
- edge detection
- prediction model
- wavelet transform
- prediction algorithm
- prosodic features
- control strategy
- transfer learning
- natural images
- prediction error
- control method
- knowledge transfer
- emotion recognition
- virtual characters
- hidden markov models
- human behaviour
- pattern recognition
- affective computing
- vocal tract
- neural network