5th International Conference on Spoken Language Processing

Sydney, Australia
November 30 - December 4, 1998

Gaussian Density Tree Structure in a Multi-Gaussian HMM-Based Speech Recognition System

Jacques Simonin, Lionel Delphin-Poulat, Geraldine Damnati

France Telecom - CNET, France

This paper presents a Gaussian density tree structure usage which enables a computational cost reduction without a significant degradation of recognition performances, during a continuous speech recognition process. The Gaussian tree structure is built from successive Gaussian density merging. Each node of the tree is associated with a Gaussian density, and the actual HMM densities are associated to the leaves. We propose then a criterion to obtain good recognition performances with this Gaussian tree structure. This structure is evaluated with a continuous speech recognition system on a telephone database. The criterion allows a 75 to 85% computational cost reduction in terms of log-likelihood computations without any significant word error rate during the recognition process.

Full Paper

Bibliographic reference.  Simonin, Jacques / Delphin-Poulat, Lionel / Damnati, Geraldine (1998): "Gaussian density tree structure in a multi-Gaussian HMM-based speech recognition system", In ICSLP-1998, paper 1063.