Hidden conditional random fields for phone classification Export

@article{citeulike:2795079, abstract = {In this paper, we show the novel application of hidden conditional random fields (HCRFs) -- conditional random fields with hidden state sequences -- for modeling speech. Hidden state sequences are critical for modeling the non-stationarity of speech signals. We show that HCRFs can easily be trained using the simple direct optimization technique of stochastic gradient descent. We present the results on the TIMIT phone classification task and show that HCRFs outperforms comparable ML and CML/MMI trained HMMs. In fact, HCRF results on this task are the best single classifier results known to us. We note that the HCRF framework is easily extensible to recognition since it is a state and label sequence modeling technique. We also note that HCRFs have the ability to handle complex features without any change in training procedure.}, author = {Gunawardana, Asela and Mahajan, Milind and Acero, Alex and Platt, John C.}, booktitle = {International Conference on Speech Communication and Technology.}, citeulike-article-id = {2795079}, citeulike-linkout-0 = {http://research.microsoft.com/research/pubs/view.aspx?type=Publication&id=1466}, keywords = {crf, discriminative, hcrf, machine-learning}, posted-at = {2008-05-13 14:56:16}, priority = {2}, title = {Hidden conditional random fields for phone classification}, url = {http://research.microsoft.com/research/pubs/view.aspx?type=Publication&id=1466}, year = {2005} }

TY - JOUR ID - citeulike:2795079 L3 - citeulike-article-id:2795079 N2 - In this paper, we show the novel application of hidden conditional random fields (HCRFs) -- conditional random fields with hidden state sequences -- for modeling speech. Hidden state sequences are critical for modeling the non-stationarity of speech signals. We show that HCRFs can easily be trained using the simple direct optimization technique of stochastic gradient descent. We present the results on the TIMIT phone classification task and show that HCRFs outperforms comparable ML and CML/MMI trained HMMs. In fact, HCRF results on this task are the best single classifier results known to us. We note that the HCRF framework is easily extensible to recognition since it is a state and label sequence modeling technique. We also note that HCRFs have the ability to handle complex features without any change in training procedure. TI - Hidden conditional random fields for phone classification T2 - International Conference on Speech Communication and Technology. KW - crf KW - discriminative KW - hcrf KW - machine-learning AU - Gunawardana, Asela AU - Mahajan, Milind AU - Acero, Alex AU - Platt, John PY - 2005/// UR - http://research.microsoft.com/research/pubs/view.aspx?type=Publication&id=1466 ER -