Prosodic Word Prediction Using a Maximum Entropy Approach.
Honghui DongJianhua TaoBo XuPublished in: ISCSLP (2006)
Keyphrases
- maximum entropy
- maximum entropy principle
- markov models
- named entity recognition
- random fields
- minimum cross entropy
- maximum entropy model
- transformation based learning
- principle of maximum entropy
- conditional random fields
- speech recognition
- class conditional
- prosodic features
- iterative scaling
- co occurrence
- prior knowledge
- generative model
- bregman divergences
- active learning