5th European Conference on Speech Communication and Technology

Rhodes, Greece
September 22-25, 1997

Fast Likelihood Computation Methods for Continuous Mixture Densities in Large Vocabulary Speech Recognition

Stefan Ortmanns, Thorsten Firzlaff, Hermann Ney

Lehrstuhl fur Informatik VI, RWTH Aachen - University of Technology, Aachen, Germany

This paper studies algorithms for reducing the computational effort of the mixture density calculations in HMM-based speech recognition systems. These likelihood calculations take about 70 total recognition time in the RWTH system for large vocabulary continuous speech recognition. To reduce the computational cost of the likelihood calculations, we investigate several space partitioning methods. A detailed comparison of these techniques is given on the North American Business Corpus (NAB'94) for a 20 000- word task. As a result, the so-called projection search algorithm in combination with the VQ method reduces the cost of likelihood computation by a factor of about 8 with no significant loss in the word recognition accuracy.

Full Paper

Bibliographic reference.  Ortmanns, Stefan / Firzlaff, Thorsten / Ney, Hermann (1997): "Fast likelihood computation methods for continuous mixture densities in large vocabulary speech recognition", In EUROSPEECH-1997, 139-142.