Speech and crosstalk detection in multichannel audio Export

@article{citeulike:4876822, abstract = {The analysis of scenarios in which a number of microphones record the activity of speakers, such as in a round-table meeting, presents a number of computational challenges. For example, if each participant wears a microphone, speech from both the microphone's wearer (local speech) and from other participants (crosstalk) is received. The recorded audio can be broadly classified in four ways: local speech, crosstalk plus local speech, crosstalk alone and silence. We describe two experiments related to the automatic classification of audio into these four classes. The first experiment attempted to optimize a set of acoustic features for use with a Gaussian mixture model (GMM) classifier. A large set of potential acoustic features were considered, some of which have been employed in previous studies. The best-performing features were found to be kurtosis, "fundamentalness," and cross-correlation metrics. The second experiment used these features to train an ergodic hidden Markov model classifier. Tests performed on a large corpus of recorded meetings show classification accuracies of up to 96\%, and automatic speech recognition performance close to that obtained using ground truth segmentation.}, author = {Wrigley, S. N. and Brown, G. J. and Wan, V. and Renals, S.}, booktitle = {Speech and Audio Processing, IEEE Transactions on}, citeulike-article-id = {4876822}, citeulike-linkout-0 = {http://dx.doi.org/10.1109/TSA.2004.838531}, citeulike-linkout-1 = {http://ieeexplore.ieee.org/xpls/abs_all.jsp?arnumber=1369314}, doi = {10.1109/TSA.2004.838531}, journal = {Speech and Audio Processing, IEEE Transactions on}, keywords = {crosstalk}, number = {1}, pages = {84--91}, posted-at = {2009-06-17 05:05:16}, priority = {2}, title = {Speech and crosstalk detection in multichannel audio}, url = {http://dx.doi.org/10.1109/TSA.2004.838531}, volume = {13}, year = {2005} }

TY - JOUR ID - citeulike:4876822 L3 - citeulike-article-id:4876822 N2 - The analysis of scenarios in which a number of microphones record the activity of speakers, such as in a round-table meeting, presents a number of computational challenges. For example, if each participant wears a microphone, speech from both the microphone's wearer (local speech) and from other participants (crosstalk) is received. The recorded audio can be broadly classified in four ways: local speech, crosstalk plus local speech, crosstalk alone and silence. We describe two experiments related to the automatic classification of audio into these four classes. The first experiment attempted to optimize a set of acoustic features for use with a Gaussian mixture model (GMM) classifier. A large set of potential acoustic features were considered, some of which have been employed in previous studies. The best-performing features were found to be kurtosis, "fundamentalness," and cross-correlation metrics. The second experiment used these features to train an ergodic hidden Markov model classifier. Tests performed on a large corpus of recorded meetings show classification accuracies of up to 96%, and automatic speech recognition performance close to that obtained using ground truth segmentation. IS - 1 JF - Speech and Audio Processing, IEEE Transactions on EP - 91 TI - Speech and crosstalk detection in multichannel audio VL - 13 T2 - Speech and Audio Processing, IEEE Transactions on SP - 84 KW - crosstalk AU - Wrigley, SN AU - Brown, GJ AU - Wan, V AU - Renals, S PY - 2005/// UR - http://dx.doi.org/10.1109/TSA.2004.838531 ER -