Text-based sentential stress prediction using continuous lexical embedding for Mandarin speech synthesis.
Yibin ZhengYa LiZhengqi WenBin LiuJianhua TaoPublished in: ISCSLP (2016)
Keyphrases
- speech synthesis
- speech recognition
- prosodic features
- text to speech
- prediction accuracy
- vocal tract
- hidden markov models
- prediction algorithm
- prediction model
- language model
- natural language processing
- pattern recognition
- vector space
- prediction error
- image search
- feature selection
- linear prediction
- data hiding
- context sensitive
- wordnet
- keywords
- semantic network
- non stationary
- sentence level
- visual features
- speaker identification
- digital images
- spoken document retrieval
- image processing
- computer vision