Feature Trajectory Dynamic Time Warping for Clustering of Speech Segments
This work addresses the challenge of accurately clustering speech segments for applications in speech processing, but it is incremental as it builds on existing dynamic time warping methods.
The authors tackled the problem of clustering speech segments by proposing a modified dynamic time warping technique that aligns feature trajectories individually, resulting in consistent and statistically significant improvements in cluster quality, with gains measured by F-measure and normalized mutual information on datasets like TIMIT and SADD.
Dynamic time warping (DTW) can be used to compute the similarity between two sequences of generally differing length. We propose a modification to DTW that performs individual and independent pairwise alignment of feature trajectories. The modified technique, termed feature trajectory dynamic time warping (FTDTW), is applied as a similarity measure in the agglomerative hierarchical clustering of speech segments. Experiments using MFCC and PLP parametrisations extracted from TIMIT and from the Spoken Arabic Digit Dataset (SADD) show consistent and statistically significant improvements in the quality of the resulting clusters in terms of F-measure and normalised mutual information (NMI).