Abstract | ||
---|---|---|
Segment model (SM) is a family of methods by using segmental distribution rather than frame-based features (e.g. HMM) to represent the underlying characteristics of observation sequence. It has been proved to be more precise than that of HMM. However, the high complexity prevents these models from practical system. In this paper we present a framework to reduce the computational complexity of segment model by fixing the number of the basic unit in the segment to share the intermediate computation results. Our work is twofold. First, we compared the complexity of SM with HMM and proposed a fast SM framework based on the comparison. Second we use two examples to illustrate this framework. The fast SMs have better performance than the system based on HMM, and at the mean time, we successfully keep the computation complexity of SM at the same level of HMM. |
Year | DOI | Venue |
---|---|---|
2004 | 10.1109/CHINSL.2004.1409600 | 2004 International Symposium on Chinese Spoken Language Processing, Proceedings |
Keywords | Field | DocType |
speech recognition,statistical distributions,computational complexity | Asymptotic computational complexity,Computer science,Speech recognition,Theoretical computer science,Probability distribution,Hidden Markov model,Computational resource,Computational complexity theory,Computation | Conference |
Citations | PageRank | References |
3 | 0.53 | 5 |
Authors | ||
4 |
Name | Order | Citations | PageRank |
---|---|---|---|
Yun Tang | 1 | 7 | 2.73 |
Wenju Liu | 2 | 214 | 39.32 |
Yiyan Zhang | 3 | 3 | 1.89 |
Bo Xu | 4 | 111 | 27.31 |