Title | ||
---|---|---|
A Maximum Entropy Based Hierarchical Model For Automatic Prosodic Boundary Labeling In Mandarin |
Abstract | ||
---|---|---|
Modeling prosodic rhythm is of great importance for both speech synthesis and speech understanding, and it requires a large enough corpus with precise prosodic boundary labels. This paper proposes a maximum entropy (ME) based hierarchical model, which utilizes both text and acoustic features, to automatically label Mandarin prosodic boundaries. Results of comparative experiments show that, for the task of prosodic boundary detection, ME model obviously outperforms classification and regression tree (CART), and the bottom-up hierarchical framework is also significantly superior to the flat single-level framework. |
Year | DOI | Venue |
---|---|---|
2008 | 10.1109/CHINSL.2008.ECP.76 | 2008 6TH INTERNATIONAL SYMPOSIUM ON CHINESE SPOKEN LANGUAGE PROCESSING, PROCEEDINGS |
Keywords | Field | DocType |
prosodic word, prosodic phrase, intonational phrase, maximum entropy | Decision tree,Speech synthesis,Pattern recognition,Computer science,Feature extraction,Speech recognition,Natural language,Artificial intelligence,Principle of maximum entropy,Hidden Markov model,Hierarchical database model,Mandarin Chinese | Conference |
Citations | PageRank | References |
16 | 1.09 | 6 |
Authors | ||
3 |
Name | Order | Citations | PageRank |
---|---|---|---|
Fangzhou Liu | 1 | 125 | 10.67 |
Huibin Jia | 2 | 19 | 2.55 |
Jianhua Tao | 3 | 848 | 138.00 |