1) HMM-based Unit Selection Speech Synthesis Using Log Likelihood Ratios Derived from Perceptual Data , Speech Communication , 2014 , vol. 63-64
2) Unsupervised Prosodic Labeling of Speech Synthesis Databases Using Context-Dependent HMMs , IEICE Transactions on Information and Systems , 2014 , vol.E97-D, no.6
3) Improving deep neural networks for LVCSR using dropout and shrinking structure , ICASSP. FLORENCE, ITALY:IEEE , 2014.5 , ICASSP-2014
4) Direct Adaptation of Hybrid DNN/HMM Model for Fast Speaker Adaptation , ICASSP. FLORENCE, ITALY:IEEE , 2014.5 , ICASSP-2014
5) An Experimental Study on Speech Enhancement Based on Deep Neural Networks , IEEE SIGNAL PROCESSING LETTERS , 2014.1 , VOL. 21, NO. 1
6) Minimum Divergence Estimation of Speaker Prior in Multi-Session PLDA , ICASSP.FLORENCE, ITALY:IEEE , 2014.5 , ICASSP-2014
7) 深度语音信号与信息处理：研究进展与展望 , 数据采集与处理 , 2014.3 , VOL.29，NO.2
8) Deep Bottleneck Features for Spoken Language Identification , PLoS One , 2014.7 , 2014 Jul 1;9(7)
9) i-vector representation based on bottleneck features for language identification , Electronics Letters , 2013 , Vol.49, no. 24
10) Minimum Kullback-Leibler Divergence Parameter Generation for HMM-based Speech Synthesis , IEEE TRANS. ON ASLP , 2012.7 , Volume:20 , Issue: 5
11) Trust Region-Based Optimization for Maximum Mutual Information Estimation of HMMs in Speech Recognition , IEEE. Trans. on ASLP , 2011 , VOL. 19, NO. 8