Fuzzy Recall and Precision for Speech Segmentation Evaluation
Bartosz Ziółko, Suresh Manandhar, Richard C. Wilson · 2007
Automatic evaluation of speech segmentation is problematic as predicted segment boundaries never align precisely. For this reason, most researchers apply ad-hoc methods for measuring the accuracy of speech segmentation. This makes judging the relative merits of each method extremely subjective and difficult. We address problem by proposing a new method for speech segmentation based on fuzzy logic. Our methods extend the well-known precision and recall metrics to the fuzzy case. The proposed method can be easily generalised to arbitrary sequence alignment and prediction problems and to arbitrary fuzzy similarity functions. 1.