Unsupervised Audio Segmentation using Extended Baum-Welch Transformations Export

@inproceedings{citeulike:1487425, abstract = {Audio segmentation has applications in a variety of contexts, such as audio information retrieval, automatic sound analysis, and as a pre-processing step in speech recognition. Extended Baum-Welch (EBW) transformations are most commonly used as a discriminative technique for estimating parameters of Gaussian mixtures. In this paper, we derive an unsupervised audio segmentation approach using these transformations. We find that our algorithm outperforms both the Bayesian information criterion (BIC) and cumulative sum (CUSUM) segmentation methods. In particular, our EBW segmentation algorithm provides improvements over the baseline approaches in detecting landmarks of short duration and minimizing landmark oversegmentation. In addition, we show that the EBW approach provides faster computation compared to the baseline methods.}, author = {Sainath, T. N. and Kanevsky, D. and Iyengar, G.}, booktitle = {Acoustics, Speech and Signal Processing, 2007. ICASSP 2007. IEEE International Conference on}, citeulike-article-id = {1487425}, citeulike-linkout-0 = {http://dx.doi.org/10.1109/ICASSP.2007.366653}, citeulike-linkout-1 = {http://ieeexplore.ieee.org/xpls/abs_all.jsp?arnumber=4217053}, doi = {10.1109/ICASSP.2007.366653}, journal = {Acoustics, Speech and Signal Processing, 2007. ICASSP 2007. IEEE International Conference on}, keywords = {algorithm, audio, segmentation}, pages = {I-209--I-212}, posted-at = {2007-11-22 01:05:46}, priority = {2}, title = {Unsupervised Audio Segmentation using Extended Baum-Welch Transformations}, url = {http://dx.doi.org/10.1109/ICASSP.2007.366653}, volume = {1}, year = {2007} }

TY - CONF ID - citeulike:1487425 L3 - citeulike-article-id:1487425 N2 - Audio segmentation has applications in a variety of contexts, such as audio information retrieval, automatic sound analysis, and as a pre-processing step in speech recognition. Extended Baum-Welch (EBW) transformations are most commonly used as a discriminative technique for estimating parameters of Gaussian mixtures. In this paper, we derive an unsupervised audio segmentation approach using these transformations. We find that our algorithm outperforms both the Bayesian information criterion (BIC) and cumulative sum (CUSUM) segmentation methods. In particular, our EBW segmentation algorithm provides improvements over the baseline approaches in detecting landmarks of short duration and minimizing landmark oversegmentation. In addition, we show that the EBW approach provides faster computation compared to the baseline methods. JF - Acoustics, Speech and Signal Processing, 2007. ICASSP 2007. IEEE International Conference on EP - I-212 TI - Unsupervised Audio Segmentation using Extended Baum-Welch Transformations VL - 1 T2 - Acoustics, Speech and Signal Processing, 2007. ICASSP 2007. IEEE International Conference on SP - I-209 KW - algorithm KW - audio KW - segmentation AU - Sainath, TN AU - Kanevsky, D AU - Iyengar, G PY - 2007/// UR - http://dx.doi.org/10.1109/ICASSP.2007.366653 ER -