Unsupervised word-level prosody tagging for controllable speech synthesis.
Yiwei GuoChenpeng DuKai YuPublished in: CoRR (2022)
Keyphrases
- speech synthesis
- word level
- speech recognition
- text to speech
- language independent
- chinese text retrieval
- prosodic features
- machine translation
- document images
- n gram
- document analysis
- part of speech
- viterbi algorithm
- word segmentation
- semi supervised
- word recognition
- character recognition
- sentence level
- metadata
- semantic roles
- novelty detection
- computer vision
- cross language information retrieval
- dynamic programming
- hidden markov models
- image processing