Transforming Spectrum and Prosody for Emotional Voice Conversion with Non-Parallel Training Data.
Kun ZhouBerrak SismanHaizhou LiPublished in: Odyssey (2020)
Keyphrases
- training data
- text to speech
- emotion recognition
- decision trees
- test data
- learning algorithm
- data sets
- synthesized speech
- real time
- supervised learning
- speech synthesis
- training set
- noisy data
- classification accuracy
- computer architecture
- language model
- information retrieval
- distributed memory
- test set
- training dataset
- training process
- parallel processing
- class labels
- shared memory
- audio visual
- domain knowledge
- parallel computation
- generalization error
- parallel programming
- cognitive radio
- classification models
- learned from training data