This paper presents some confidence measures for large vocabulary speech recognition which are based on word graph or N-Best List structures. More and more applications need fast estimation of any measures in order to stay real-time. We propose some simple and fast measures, locally computed, that can be directly used within the first decoding recognition process. We also define some other measures combining word graph and N-Best List. Experimental results on a French broadcast news corpus are also presented.
Cite as: Razik, J., Mella, O., Fohr, D., Haton, J.-P. (2005) Local word confidence measure using word graph and n-best list. Proc. Interspeech 2005, 3369-3372, doi: 10.21437/Interspeech.2005-865
@inproceedings{razik05_interspeech,
author={Joseph Razik and Odile Mella and Dominique Fohr and Jean-Paul Haton},
title={{Local word confidence measure using word graph and n-best list}},
year=2005,
booktitle={Proc. Interspeech 2005},
pages={3369--3372},
doi={10.21437/Interspeech.2005-865}
}