Issue No. 09 - September (2007 vol. 56)
DOI Bookmark: http://doi.ieeecomputersociety.org/10.1109/TC.2007.70746
Jia-Ching Wang , IEEE
Jhing-Fa Wang , IEEE
<p><b>Abstract</b>—This work presents an unsupervised speaker change detection algorithm based on support vector machine (SVM) to detect speaker change in a speech stream. The proposed algorithm is called the SVM training misclassification rate (STMR). The STMR can identify speaker changes with less speech data collection, making it capable of detecting speaker segments with short duration. According to experiments on the NIST Rich Transcription 2005 Spring Evaluation (RT-05S) corpus, the STMR has a missed detection rate of only 19.67%.</p>
Speaker segmentation, Support Vector Machine, Speaker Change Detection
P. Lin, J. Wang, H. Sung and J. Wang, "Unsupervised Speaker Change Detection Using SVM Training Misclassification Rate," in IEEE Transactions on Computers, vol. 56, no. , pp. 1234-1244, 2007.