夏树涛教授、博士生导师 电话：　　
邮箱： xiast@sz.tsinghua.edu.cn办公地址：南山区西丽大学城清华园区H栋202A
所属学科：计算机科学与技术
职务：深圳研究生院计算机科学与技术研究所所长
深圳智能语义挖掘技术工程实验室主任


【个人简历】
1988年9月-1992年7月，南开大学数学专业本科毕业，获学士学位
1992年9月-1997年1月，南开大学信息论与编码专业毕业，获博士学位
1997年9月-1998年9月，香港中文大学信息工程系访问
2004年1月至今，清华大学深圳研究生院先后任副研究员、教授（2007）、博士生导师（2009）
【教学】
研究生课程：《信道编码》、《应用信息论基础》
【研究领域】
主要从事信息论编码、互联网与人工智能等方向的教学与科研工作，目前的主要研究兴趣为编码/量化、AI互联网及机器学习算法等。在IEEE Transactions on Information Theory (TIT), TSP, TNNLS,TCOM,TCSVT, Network等国际权威期刊和NIPS, ICML, CVPR, IJCAI, AAAI, DCC, ISIT, ICASSP等领域顶级会议上发表论文六十多篇。近年来完成多项国家级科研项目，其中主持的项目有：
1．复杂动态互联网行为的基准建模与异常分析，国家重点研发计划重点专项，课题负责人，2019-2022.
2.未来互联网的高效路由与智能传输机理，国家973课题，2012.1-2016.12，527万.
3．基于校验矩阵方法的局部修复码研究与应用，国家自然科学基金，2018.1-2021.12，67万.
4．基于LDPC码的压缩感知测量矩阵构造及性能分析，国家自然科学基金，2014.1-2017.12，82万.
5．Non-coherent网络中的纠错码及其应用，国家自然科学基金，2010.1-2012.12，30万.
6．LDPC码的译码性能分析及应用，国家自然基金委与广东省政府联合基金，2007.1-2009.12，30万.
7．二元码的检错/纠错性能估计和应用，国家自然科学基金，2005.1-2007.12，23万.
8. 面向未来网络的内容分发技术研究，深圳市科创委技术攻关项目，2015.7-2017.12，400万.
9．深圳智能语义挖掘技术工程实验室，深圳市发改委，2012.11- 2014.12，500万.
10.大规模机器学习算法的噪声鲁棒性及在证券交易领域的应用研究，深圳市基础研究学科布局项目，2019.1-2021.12，300万.
【论文与专利】



机器学习、稀疏表示及应用
[1]Tao Dai, Li Niu, and Shu-Tao Xia. Second-order attention network for single image super-resolution, Proc. IEEE Conference on Computer Vision and Pattern Recognition(CVPR-19), oralpaper,Long Beach, CA, U.S.A., Jun. 2019.
[2]Xiyu Yan, Shuai Chen, Zihao He, Chunmei Li, Feng Zheng, Tao Dai, Shuo Dong, Yong Jiang, Shu-Tao Xia. Automatic grassland degradation estimation using deep learning, Proc. the 28th International Joint Conference on Artificial Intelligence(IJCAI-19), Macao, China, Aug. 2019.
[3]Xi Xiao, Rui Li, Hai-Tao Zheng, Runguo Ye, Arun KumarSangaiah, Shutao Xia, Novel dynamic multiple classification system for network traffic, Information Sciences (IS), vol. 479, pp. 526-541, 2019.
[4]Jin-Yuan Chen, Hai-Tao Zheng, Yong Jiang, Shu-Tao Xia, Cong-Zhi Zhao. A probabilistic model for semantic advertising, Knowledge and Information Systems(KIS), vol. 59, pp. 387-412, May 2019.
[5]Weizhi Lu, Weiyu Li, Wei Zhang, and Shu-Tao Xia. Expander recovery performance of bipartite graphs with girth greater than 4, IEEE Transactions on Signal and Information Processing over Networks (TSIPN), accepted,online published via https:// ieeexplore.ieee.org/document/**.
[6]Songtao Wang, Dan Li, Yang Cheng, Jinkun Geng, Yanshu Wang, Shu-Tao Xiaand Jianping Wu. BML: a high-performance, low-cost gradient synchronization algorithm for DML training,Proc. Neural Information Processing Systems(NIPS-18), Montreal, Canada, Dec. 2018.
[7]Xingjun Ma, Yisen Wang, Michael E. Houle, Shu-Tao Xia, James Bailey.Dimensionality-driven learning with noisy labels, Proc. the Thirty-fifth International Conference on Machine Learning(ICML-18), long talk, Stockholm, Sweden, Jul. 2018.
[8]Yisen Wang, Weiyang Liu, Xingjun Ma, James Bailey, Hongyuan Zha, Le Song,and Shu-Tao Xia. Iterative learning with open-set noisy labels, Proc. IEEE Conference on Computer Vision and Pattern Recognition(CVPR-18), spotlight paper,Salt Lake City, U.S.A., Jun. 2018.
[9]Yisen Wang, Shu-Tao Xia, Qingtao Tang, Jia Wu, Xingquan Zhu. A novel consistent random forests framework: Bernoulli random forests.IEEE Transactions on Neural Networks and Learning Systems(TNNLS), vol. 29, no. 8, pp. 3510-3523, Aug. 2018.
[10]Rui Li, Xi Xiao, Shiguang Ni, Haitao Zheng, Shutao Xia, Byte segment neural network for network traffic classification, Proc. IEEE/ACM International Symposium on Quality of Service(IWQoS-18), Banff, Alberta, Canada, Jun. 2018.
[11]Weizhi Lu, Tao Dai, Shu-Tao Xia. Binary matrices for Compressed Sensing, IEEE Transactions on Signal Processing (TSP), vol. 66, no. 1, pp. 77-85, Jan. 2018.
[12]Yang Zhang, Qingtao Tang, Li Niu, Tao Dai, Xi Xiao, Shu-Tao Xia. Self-paced mixture of t distribution model, Proc.the 43nd IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP-18), Calgary, Canada, Apr. 2018.
[13]Zhiya Xu, Tao Dai, Li Niu, Jiawei Li, Qingtao Tang, Shu-Tao Xia. Sure-based dual domain image denoising, Proc.the 43nd IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP-18), Calgary, Canada, Apr. 2018.
[14]Chaobing Song, Shaobo Cui, Yong Jiang, Shu-Tao Xia. Accelerated stochastic greedy coordinate descent by soft thresholdingprojection onto simplex. Proc. Neural Information Processing Systems(NIPS-17), spotlight paper, Long Beach, U.S.A., Dec. 2017.
[15]Qingtao Tang, Tao Dai, Li Niu, Yisen Wang, Shu-Tao Xia, and Jianfei Cai.Robust Survey Aggregation with Student-t Distribution and Sparse Representation. Proc. the 26th International Joint Conference on Artificial Intelligence(IJCAI-17), pp. 2829-2835, Melbourne, Australia, Aug. 2017.
[16]Qingtao Tang, Li Niu, Yisen Wang, Tao Dai, Wangpeng An, Jianfei Cai, Shu-Tao Xia. Student-t Process Regression with Student-t Likelihood. Proc. the 26th International Joint Conference on Artificial Intelligence(IJCAI-17), pp. 2822-2828, Melbourne, Australia, Aug. 2017.
[17]Xin-Ji Liu, Shu-Tao Xia, and Fang-Wei Fu, Reconstruction guarantee analysis of basis pursuitfor binary measurement matrices in compressedsensing, IEEE Transactions on Information Theory (TIT), vol. 63, no. 5, pp. 2922-2932,May 2017.
[18]Y. Wang, S. Romano, N.X. Vinh, J. Bailey, X. Ma, S.-T. Xia. Unbiased multivariate correlation analysis, Proc. the 31st AAAI Conference on Artificial Intelligence(AAAI-17),San Francisco, California, U.S.A., Feb. 2017.
[19]Yisen Wang, Fangbing Liu, Shu-Tao Xia, Jia Wu.Link sign prediction by variational Bayesian probabilistic matrix factorization with student-t prior, Information Sciences(IS), vol. 405, pp. 175-189, 2017.
[20]Yisen Wang, Shu-Tao Xia, and Jia Wu, A less-greedy two-term Tsallis Entropy Information Metric approach for decision tree classification, Knowledge-Based Systems (KBS), vol.120, pp. 34-42, Mar. 2017.
[21]Yisen Wang and Shu-Tao Xia, Unifying attribute splitting criteria of decision trees by Tsallis entropy, Proc.the 42nd IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP-17), pp. 2507-2511, New Orleans, U.S.A., Mar. 2017.
[22]Weizhi Lu, Tao Dai, and Shu-Tao Xia, Compressed Sensing Performance of Binary Matrices with Binary Column Correlations, Proc. 2017 Data Compression Conference(DCC-17), pp. 151-160, Snowbird, Utah, U.S.A., Apr. 2017.
[23]YisenWang, Qingtao Tang, Shu-Tao Xia, Jia Wu, Xingquan Zhu, Bernoulli random forests: closing the gap between theoretical consistency and empirical soundness,Proc. the 25th International Joint Conference on Artificial Intelligence(IJCAI-16), pp. 2167-2173, New York, U.S.A., Jul. 2016.
[24]Qingtao Tang, Yisen Wang, and Shu-Tao Xia, Student-t process regression with dependent student-t noise, Proc. the 22nd EuropeanConference onArtificialIntelligence(ECAI-16), pp. 82-89, Hague, Holland, Aug. 29 - Sep. 2, 2016.
[25]Tao Dai, Ke Gu, Li Niu, Yong-bing Zhang, Weizhi Lu, Shu-Tao Xia.Referenceless quality metric of multiply-distorted images based on structural degradation. Neurocomputing, vol. 290, pp. 185-195, May 2018.
[26]Tao Dai, Zhiya Xu,Haoyi Liang,Ke Gu,Qingtao Tang,Yisen Wang,Weizhi Lu,Shu-Tao Xia. A generic denoising framework via guided principal component analysis,Journal of Visual Communication and Image Representation, vol. 48, pp. 340-352, Oct.2017.
[27]Tao Dai, Weizhi Lu, Wei Wang, Jilei Wang, and Shu-Tao Xia, Entropy-based bilateral filtering with a new range kernel, Signal Processing, vol. 137, pp. 223-234,Aug. 2017.
[28]Weizhi Lu and Shu-Tao Xia, Construction of ternary matrices with smallcoherence for compressed sensing, Electronics Letters(EL), vol. 52, no. 6,pp. 447–448, Mar.2016.
[29]S.-T. Xia, X.-J. Liu, Y. Jiang,and H.-T. Zheng, Deterministic constructions of binary measurement matrices from finite geometry, IEEE Transactions on Signal Processing (TSP), vol. 63, no. 4, pp. 1017-1029, Feb. 2015.
[30]X.-J. Liu, S.-T. Xia, T. Dai, Deterministic construction of binary measurementmatrices with various sizes, Proc.40th IEEE International Conference onAcoustics, Speech and Signal Processing (ICASSP-15), pp. 3641-3645, Brisbane, Australia, Apr.2015.
[31]C.-B. Song, S.-T. Xia, X.-J. Liu, Subspace thresholding pursuit: A reconstruction algorithm for compressed sensing, Proc. IEEE International Symposium on Information Theory (ISIT-15), pp. 536-540, Hong Kong, China, Jun. 2015.
[32]C.-B. Song and S.-T. Xia, Sparse signal recovery by l_q minimization under restricted isometry property, IEEE Signal Processing Letters (SPL), vol. 21, no. 9, pp. 1154-1158, Sep. 2014.
[33]C.-B. Song, S.-T. Xia, and X.-J. Liu, Improved analysis for subspace pursuit algorithmin terms of restricted isometry constant, IEEE Signal Processing Letters (SPL), vol. 21, no. 11, pp. 1365-1369, Nov. 2014.
[34]F. Zhou, S.-T. Xia, and Q. Liao. Nonlocal pixel selection for multisurface fitting-based super-resolution. IEEE Transactions on Circuitsandsystems for video technology(TCSVT), vol. 24, no.12, pp. 2013-2017, Dec. 2014.
信息论编码及网络
[35]Bin Chen, Weijun Fang, Shu-Tao Xia, and Fang-Wei Fu. Constructions of optimal (r,δ) locally repairable codes via constacyclic Codes, IEEE Transactions on Communications (TCOM) , accepted, May 2019, online published via https:// ieeexplore.ieee.org/document/**.
[36]Bin Chen, Shu-Tao Xia, Jie Hao, and Fang-Wei Fu. Constructions of optimal cyclic (r,δ) locally repairable codes, IEEE Transactions on Information Theory (TIT), vol. 64, no. 4, pp. 2499-2511, Apr. 2018.
[37]Xi Xiao, Peng Fu, Changsheng Dou, Qing Li, Guangwu Hu, and Shutao Xia. Design and analysis of SEIQR worm propagation model in mobile internet, Communications in Nonlinear Science and Numerical Simulation,vol.43,pp.341-350,Feb.2017.
[38]Qing Li, Mingwei Xu, Qi Li, Dan Wang, Yong Jiang, Shu-Tao Xia, Qingmin Liao.Scale the Internet routing table by generalized next hops of strict partial order, Information Sciences(IS), vol. 412, pp. 101-115, Oct. 2017.
[39]Yong Cui,Lian Wang,Xin Wang,Yisen Wang,Fengyuan Ren,Shutao Xia.End-to-end coding forTCP, IEEE Network, vol. 30, no. 2, pp. 68-73, Mar. 2016.
[40]Y. Jiang, S.-T. Xia,and F.-W. Fu. Stoppingset distributions of some kinds of Reed-Muller codes, IEEE Transactions on Information Theory (TIT), vol. 57, no. 9, pp. 6078-6088, Sep. 2011.
[41]S.-T. Xiaand F.-W. Fu, Minimum pseudo-weight and minimum pseudo-codewords of LDPC Codes, IEEE Transactions on Information Theory (TIT), vol. 54, no. 1, pp. 480-485, Jan. 2008.
[42]S.-T. Xia, F.-W. Fu, and S. Ling, A lower bound on the probability of undetected errorfor binary constant weight codes. IEEE Transactions on Information Theory (TIT),vol. 52, no. 9, pp. 4235-4243, Sep. 2006.
[43]S.-T. Xia, F.-W. Fu, Y. Jiang, and S. Ling, The probability of undetected error for binary constant weight codes. IEEE Transactions on Information Theory (TIT),vol. 51, no. 9, pp. 3364-3373,Sep.2005.
[44]F.-W. Fu, T. Klove,and S.-T. Xia, The undetected error probability threshold ofm-out-of-n codes, IEEE Transactions on Information Theory (TIT),vol.46, no.4,pp.1597-1599, Jul. 2000.
[45]F.-W. FuandS.-T. Xia,Binary constant weight codes for error detection, IEEE Transactions on Information Theory (TIT), vol. 44, no. 3, pp. 1294-1299, May 1998.
[46]S.-T. Xiaand F.-W. Fu, Johnson type bounds on constant dimension codes. Designs, Codes and Cryptography,vol.50,no.2,pp. 163-172, 2009.
[47]S.-T. Xiaand F.-W. Fu, Undetected error probability of q-ary constant weight codes. Designs, Codes and Cryptography,vol.48, pp. 125-140, 2008.
[48]F.-W. Fuand S.-T. Xia,The characterization of binary constant weight codes meeting thebound of Fu and Shen, Designs, Codes and Cryptography,vol.43, pp. 9-20, 2007.
[49]S.-T. Xia, F.-W. Fu, and Y. Jiang, On the minimum average distance of binary constant weight codes, Discrete Mathematics,vol. 308, pp. 3847-3859, 2008.
[50]S.-T. Xiaand F.-W. Fu, On the average Hamming distance for binary codes. Discrete Applied Mathematics, vol. 89, pp. 269-276, 1998.
[51]J.Hao and S.-T. Xia, Constructions of optimal binary locally repairable codes with multiple repair groups, IEEE Communications Letters, vol.20, no.6, pp.1060-1063, Jun. 2016.
[52]X. Ge, S.-T. Xia,Structured non-binary LDPC codes with large girth. Electronics Letters, vol. 43, no. 22, pp: 1220-1222, Oct. 25, 2007
[53]S.-T. Xiaand F.-W. Fu, On the minimum pseudo-codewords of LDPCcodes, IEEE Communications Letters, vol. 10, no. 5, pp.363-365, May 2006.
[54]S.-T. Xiaand F.-W. Fu, On the stopping distance of finitegeometry LDPC Codes, IEEE Communications Letters, vol.10, no. 5, pp. 381-383, May 2006.
[55]Jie Hao, Kenneth W. Shum, Shu-Tao Xia, Yi-Xian Yang. Classification of optimal ternary (r,\delta)-locally repairable codes attaining the Singleton-like bound, Proc. IEEE International Symposium on Information Theory(ISIT-19), accepted, Vail, Paris, France, Jul. 2019.
[56]Bin Chen, Shu-Tao Xia, Jie Hao. Improved bounds and optimal constructions of locally repairable codes with distance 5 and 6, Proc. IEEE International Symposium on Information Theory(ISIT-19), accepted, Vail, Paris, France, Jul. 2019.
[57]Bin Chen, Shu-Tao Xia, Jie Hao, and Fang-Wei Fu, On optimal pseudo-cyclic (r,δ) locally repairable codes, Proc. IEEE International Symposium on Information Theory(ISIT-18), pp. 1191-1195, Vail, Colorado, USA, Jun. 2018.
[58]Jie Hao, Kenneth W. Shum, Shu-Tao Xia, and Yixian Yang, On the maximal code length of optimal linear locally repairable codes, Proc. IEEE International Symposium on Information Theory(ISIT-18), pp. 1326-1330, Vail, Colorado, USA, Jun. 2018.
[59]Jie Hao, Shu-Tao Xia, and Bin Chen, On optimal ternary locally repairable codes, Proc. IEEE International Symposium on Information Theory(ISIT-17), pp. 171-175, Aachen, Germany, Jun. 2017.
[60]Bin Chen, Shu-Tao Xia, and Jie Hao, Locally repairable codes with multiple (ri, δi)-localities, Proc. IEEE International Symposium on Information Theory(ISIT-17), pp. 2038-2042, Aachen, Germany, Jun. 2017.
[61]Jie Hao, Shu-Tao Xia, and Bin Chen.On the linear codes with (r,δ)-locality for distributed storage, Proc. IEEE International Conference on Communications(ICC-17), pp. 1-6, Paris, French, May 2017.
[62]J. Hao, S.-T. Xia,and B. Chen, Some results on optimal locally repairable codes, Proc. IEEE International Symposium on Information Theory(ISIT-16), pp. 440-444, Barcelona, Spain, Jul. 2016.
[63]J. Hao, S.-T. Xia,and B. Chen, Recursive bounds for locally repairable codes with multiple repair groups, Proc. IEEE International Symposium on Information Theory(ISIT-16), pp. 645-649, Barcelona, Spain, Jul. 2016.
[64]X.-J. Liu, S.-T. Xia, Reconstruction guarantee analysis of binary measurement matrices based on girth, Proc. IEEE Int. Sym. Inform. Theory (ISIT-13), pp. 474-478, Istanbul, Turkey, Jul. 2013.
[65]X.-J. Liu, S.-T. Xia, Constructions of quasi-cyclic measurement matrices based on array codes, Proc. IEEE Int. Sym. Inform. Theory(ISIT-13), pp. 479-483, Istanbul, Turkey, Jul. 2013.
[66]Yong Jiang, Shu-Tao Xia, Xin-Ji Liu, and Fang-Wei Fu. Incorrigible set distributions and unsuccessful decoding probability of linear codes, Proc. IEEE International Symposium on Information Theory(ISIT-13), pp. 1042-1046, Istanbul, Turkey, Jul. 2013.
【兼职情况】

中国电子学会高级会员
中国电子学会信息论分会理事