ISCA Archive Eurospeech 1999
ISCA Archive Eurospeech 1999

Task dependent loss functions in speech recognition: a* search over recognition lattices

Vaibhava Goel, William Byrne

A recognition strategy that can be matched to specific system performance criteria such as word error rate or F-measure has recently been found to yield improvements over the usual maximum aposteriori probability strategy [1] [2] [3]. In this matched-to-the-task strategy a hypothesis is chosen to minimize the expected loss or the Bayes Risk under a loss function defined by a performance measure of interest. Due to the prohibitive of exact implementation of this strategy, only an approximate implementation as an N-best list rescoring scheme been used [1] [2]. Our goal is to improve the performance of such risk-based decoders by developing search strategies that can consider more hypotheses and incorporate more acoustic evidence. In this paper we present search algorithms to implement the risk-based recognition strategy over word lattices that contain acoustic and language model scores. These algorithms are extensions of the N-best list rescoring approximation and are formulated as A* algorithms. Results are reported on the Switch-board conversational telephone speech corpus. We find that lattice based rescoring yields modest but significant improvements in word error rate relative to N-best list rescoring at comparable computational cost.


doi: 10.21437/Eurospeech.1999-291

Cite as: Goel, V., Byrne, W. (1999) Task dependent loss functions in speech recognition: a* search over recognition lattices. Proc. 6th European Conference on Speech Communication and Technology (Eurospeech 1999), 1243-1246, doi: 10.21437/Eurospeech.1999-291

@inproceedings{goel99_eurospeech,
  author={Vaibhava Goel and William Byrne},
  title={{Task dependent loss functions in speech recognition: a* search over recognition lattices}},
  year=1999,
  booktitle={Proc. 6th European Conference on Speech Communication and Technology (Eurospeech 1999)},
  pages={1243--1246},
  doi={10.21437/Eurospeech.1999-291}
}