5th European Conference on Speech Communication and Technology

Rhodes, Greece
September 22-25, 1997

Maximum Likelihood Successive State Splitting Algorithm for Tied-Mixture HMNET

Alexandre Girardi, Harald Singer, Kiyohiro Shikano, Satoshi Nakamura

Nara Institute of Science and Technology, Ikoma-shi, Nara-ken, Japan

This paper describes a new approach to ML-SSS (Maximum Likelihood Successive State Splitting) algorithm that uses tied- mixture representation of the output probability density function instead of a single Gaussian during the splitting phase of the ML-SSS algorithm. The tied-mixture representation results in a better state split gain, because it is able to measure diferences in the phoneme environment space that ML-SSS can not. With this more informative gain the new algorithm can choose a better split state and corresponding data. Phoneme clustering experiments were conducted which lead up to 38% of error reduction if compared to the ML-SSS algorithm.

Full Paper

Bibliographic reference.  Girardi, Alexandre / Singer, Harald / Shikano, Kiyohiro / Nakamura, Satoshi (1997): "Maximum likelihood successive state splitting algorithm for tied-mixture HMNET", In EUROSPEECH-1997, 119-122.