Automatic prosody prediction for Chinese speech synthesis using BLSTM-RNN and embedding features.
Chuang DingLei XieJie YanWeini ZhangYang LiuPublished in: ASRU (2015)
Keyphrases
- speech synthesis
- speech recognition
- text to speech
- prosodic features
- vocal tract
- feature vectors
- prediction accuracy
- feature set
- nearest neighbor
- recurrent neural networks
- feature space
- feature extraction
- image processing
- co occurrence
- image features
- image quality
- splice site
- prediction algorithm
- extracted features
- neural network
- training data
- machine learning