5th International Conference on Spoken Language Processing

Sydney, Australia
November 30 - December 4, 1998

A Duration-Based Confidence Measure for Automatic Segmentation of Noise Corrupted Speech

Bryan L. Pellom, John H. L. Hansen

Duke University, USA

In this study, a duration-based measure is formulated for assigning confidence scores to phonetic time-alignments produced by an automatic speech segmentation system. For speech corrupted by additive noise or telephone channel environments, the proposed confidence measure is shown to provide a reliable means by which gross segmentation errors can be automatically detected and marked for human hand correction. The measure is evaluated by computing Receiver Operating Characteristic (ROC) curves to illustrate the expected trade-off in probability of detecting gross segmentation errors versus false alarm rates.

Full Paper

Bibliographic reference.  Pellom, Bryan L. / Hansen, John H. L. (1998): "A duration-based confidence measure for automatic segmentation of noise corrupted speech", In ICSLP-1998, paper 0853.