基本信息

侯新文  男  硕导  中国科学院自动化研究所
电子邮件: xinwen.hou@ia.ac.cn
通信地址: 海淀区中关村东路95号
邮政编码: 100190

研究领域

模式识别、计算机视觉、深度学习、强化学习、生成对抗学习、智能博弈、人工智能的数学理论

研究经历

侯新文,2001年毕业于北京大学数学系,获理学博士学位,同年进入南开大学数学系攻读博士后,出站后任中国科学院自动化研究所模式识别国家重点实验室副研究员,现为中科院自动化所项目研究员、硕士研究生导师。本人长期从事计算机视觉和机器学习研究,在人脸识别领域提出了直接形象模型(DAM) ,在子空间学习领域提出局部化非负矩阵分解(LNMF),在图像文字检测中提出使用条件虽机场过滤文本部件,在物体检测、运动跟踪、流形学习、AdaBoost、机器学习基础理论等方面发表过多篇文章,Google Scholar 2200多次 。近年来工作重点转向强化学习、生成对抗学习、元学习、智能博弈等通用人工智能方向,以在游戏博弈中战胜人类选手为应用目标,以提出人工智能的统一数学理论为长期研究目标。



高引用论文

Google Scholar引用2796多次,其中超过100次引用的论文如下:

[1]  Learning Spatially Localized, Parts-Based Representation, 927次

[2]  A hybrid approach to detect and localize texts in natural scene images, 367次

[3]  Direct Appearance Models,219次

[4]  Text localization in natural scene images based on conditional random field,144次

[5]  Learning Multiview Face Subspaces and Facial Pose Estimation Using Independent Component Analysis,112次


arxiv预印论文

[1] Pei Yingjun, Hou Xinwen, Learning Representations in Reinforcement Learning: An Information Bottleneck Approach.https://arxiv.org/abs/1911.05695

[2] Chen Gong, Qiang He, Yunpeng Bai, Xiaoyu Chen, Xinwen Hou, Yu Liu, Guoliang Fan, Combing Policy Evaluation and Policy Improvement in a Unified f-Divergence Framework. https://arxiv.org/abs/2109.11867

[3] Xiaoyu Chen, Chen Gong, Qiang He, Xinwen Hou, Yu Liu, LDC-VAE: A Latent Distribution Consistency Approach to Variational AutoEncoders. https://arxiv.org/abs/2109.10640


招生信息

2023年度招收1名全日制硕士和2名非全日制硕士,研究深度学习、生成对抗网络、计算机视觉或强化学习的理论方法和应用技术

招生专业
081104-模式识别与智能系统
招生方向
计算机视觉
模式识别理论与方法
强化学习

教育背景

1998-09--2001-06   北京大学数学系   博士

工作经历

   
工作简历
2004-07~现在, 中国科学院自动化研究所, 副研究员
2003-12~2004-06,中科院自动化所, 项目聘用
2001-07~2003-11,南开大学数学系, 博士后

教授课程

模式识别课

出版信息

   
期刊论文

[17] 范波,钟季龙,徐丽霞,吕筱璇,王鹥喆,刘禹,侯新文,基于因果熵的无人集群对抗评估指标分配方法,系统工程与电子技术,2023.

[16] Chenhao Li, Yuhui Fu, Ruihong Ouyang, Yu Liu, Xinwen Hou. ADTIDO: Detecting the Tired Deck Officer with Fusion Feature Methods[J]. Sensors, 2022, 22(17): 6506.

[15] Jiangning Wang, Yingying Chen,  Xinwen Hou., Yong Wang, Libing Zhou, Xiaolin Chen, An intelligent identification system combining image and DNA sequence methods for fruit flies with economic importance (Diptera: Tephritidae). Pest Manag. Sci. 77(7): 3382--3395,2021.

[14] 申翔翔,侯新文,尹传环,深度强化学习中状态注意力机制的研究,智能系统学报,2019.

[13] D. Wang, X. W. Hou, J. Xu, S. Yue, C.-L. Liu, Traffic sign detection using a cascade method with fast feature extraction and saliency test, IEEE Trans. Intelligent Transportation Systems, 18(12): 3290-3302, 2017.

[12] Jiang-ning Wang, Xiao-lin Chen, Xin-wen Hou, Li-bing Zhou, Chao-Dong Zhu and Li-qiang Ji, Construction, implementation and testing of an image identification system using computer vision methods for fruit flies with economic importance (Diptera: Tephritidae), Pest Management Science, Wiley Online Library, 73(7):1511-1528, 2016.

[11] Y.-M. Zhang, K. Huang, X. Hou, C.-L. Liu, Learning locality preserving graph from data, IEEE Trans. SMC Part B, 44(11): 2088-2098, 2014.

[10] Guoqiang Zhong, Kaizhu Huang, Xinwen Hou, S. Xiang, Local Tangent Space Laplacian Eigenmaps, Computational Intelligence, 2012

[9] Y.-F. Pan, X. W. Hou, C.-L. Liu, A hybrid approach to detect and localize texts in natural scene images, IEEE Trans. Image Processing, 20(3): 800-813, 2011.

[8] 张蕾,陈小琳,侯新文,刘成林,樊利民,汪兴鉴,实蝇科果实蝇属昆虫数字图像自动识别系统的构建和测试,昆虫学报,54(2):184-196,2011.

[7]  X.-B. Jin, C.-L. Liu, X. W. Hou, Regularized margin-based conditional log-likelihood loss for prototype learning, Pattern Recognition, 43(7): 2428-2438, 2010.

[6]  陈小琳,侯新文,刘成林,刘晓秋,张知彬,昆虫图像自动鉴别技术,昆虫知识,45(2): 317-322, 2008.

[5]  W. Dong, X. Hou, J. Liu,Y. Fang, C. Jin and Q. Zhu, 3D virtual reconstruction of the pleistocene cheetah skull  from the Tangshan, Nanjing, China, Progress In Natural Science, vol. 17(1),  pp. 74-79, 2007.

[4]  董为,侯新文,房迎三,刘金毅,朱奇志, 南京汤山早更新世猎豹头骨CT扫描数据的三维重建,自然科学进展,vol. 16(4), pp. 1146-1152, 2006.

[3]  Stan Z.  Li, X. G. Lv, X. W. Hou, X. H. Peng and Q. S. Cheng, Learning Multiview Face Subspaces and Facial Pose Estimation

Using Independent Component Analysis, IEEE Trans. Image Processing, 14(6):705- 712, 2005. 

[2]  Shuicheng Yan, Xinwen Hou, Stan Z. Li, Hongjiang Zhang, and Qiansheng Cheng, Face Alignment Using View-Based Direct Appearance Models, Special Issue on Facial Image Processing, Analysis and Synthesis, International Journal of Imaging Systems and Technology, Vol.1, p106-112, 2003.

[1]  侯新文,程乾生,一种改进的变分Snake模型,数学的理论与实践,vol. 31(2), pp. 202-205, 2001.



专著章节

[1] A. Lu, X. W. Hou, C.-L. Liu, X. Chen, Insect recognition using sparse coding and decision fusion, In: Computer Vision and Pattern Recognition in Environmental Informatics, Jun Zhou, Xiao Bai and Terry Caelli (Eds.), IGI Global, pp.124-145, 2015.

会议论文

[53] Qiang He, Xinwen Hou, MEPE: A Minimalist Ensemble Policy Evaluation Operator For Deep Reinforcement Learning, The IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP), 2024. 

[52] Chao Li, Chen Gong, Qiang He, and Xinwen Hou. Keep Various Trajectories: Promoting Exploration of Ensemble Policies in Continuous Control, Thirty-seventh Conference on Neural Information Processing Systems (NeurIPS), 2023.

[51] Bo Fan, Xiaoxuan Lv, Jilong Zhong, Lixia Xu, Shaoshi Wu, Yishan Ding, Xiaoyu Zhai, and Xinwen Hou,Unmanned Cluster Intelligence Level Evaluation Based on ELO Scores and Generative Adversarial Network, The International Conference on Information Technology, Computing Science, and Intelligent Systems (ITCSIS), 2023. 

[50] Hewei Guo, Bo Fan, Jilong Zhong, Lixia Xu, Shaoshi Wu, Shaoshi Wu, Xiaoyu Zhai, Xinwen Hou,  Unsupervised Domain Adaptation via Adversarial Attack Consistency, China Automation Congress (CAC),  2023.

[49] Jiaqi Wang, Bo Fan, Jilong Zhong, Lixia Xu, Xiaoyu Zhai, Le Zhou, Yu Liu, Xinwen Hou, An Efficient Metamorphic Testing Leveraging Disentangled Transformation and Decision Oracle, IEEE 6th International Conference on Pattern Recognition and Artificial Intelligence(PRAI), 2023.

[48] Hewei Guo, Liping Ren, Jingjing Fu, Yuwang Wang, Zhizheng Zhang, Cuiling Lan, Haoqian Wang, Xinwen Hou, Template-guided Hierarchical Feature Restoration for Anomaly Detection,  International Conference on Computer Vision (ICCV), 2023.

[47] Chao Li, Chen Gong, Xinwen Hou, Yu Liu, Qiang He. Centralized Cooperative Exploration Policy for Continuous Control Tasks,  International Joint Conference on Autonomous Agents and Multi-agent Systems (AAMAS), 2023.

[46]Qiang He, Huangyuan Su, Jieyu Zhang,  Xinwen Hou,  Frustratingly Easy Regularization on Representation Can Boost Deep Reinforcement Learning, The IEEE Conference on Computer Vision and Pattern Recognition(CVPR),2023.

[45]Chen Gong, Zhou Yang, Yunpeng Bai, Jieke Shi, Arunesh Sinha, Bowen Xu, David Lo, Xinwen Hou, Guoliang Fan, Curiosity-Driven and Victim-Aware Adversarial Policies, Annual Computer Security Applications Conference (ACSAC), 2022.

[44]Yunpeng Bai, Chen Gong, Bin Zhang, Guoliang Fan, Xinwen Hou, Yu Liu, Cooperative Multi-Agent Reinforcement Learning with Hypergraph Convolution, International Joint Conference on Neural Networks(IJCNN), IEEE WCCI 2022.

[43]Qiang He, Huangyuan Su, Jieyu Zhang, Xinwen Hou,  Representation Gap in Deep Reinforcement Learning, Decision Awareness in Reinforcement Learning workshop at the International Conference on Machine Learning (ICML) 2022.

[42] Qiang He, Huangyuan Su, Chen Gong, Xinwen Hou,  MEPG: A Minimalist Ensemble Policy Gradient Framework for Deep Reinforcement Learning, Decision Awareness in Reinforcement Learning workshop at the International Conference on Machine Learning (ICML) 2022.

[41] Qiang He, Xinwen Hou, Yu Liu,  POPO: Pessimistic Offline Policy Optimizaiton,IEEE International Conference on Acoustics, Speech and Signal Processing(ICASSP), 2022.

[40] Yingying Chen, Xinwen Hou, and Yu Liu, Minimizing Wasserstein-1 Distance by Quantile Regression for GANs Model,The Fourth China pattern recognition and computer vision Conference (PRCV), 2021.

[39] Chen Gong, Qiang He, Yunpeng Bai, Xinwen Hou, Guoliang Fan, Yu Liu, Wide-sense Stationary Policy Optimization with Bellman Residual on Video Games, International Conference on Multimedia and Expo (ICME), 2021.

[38] Qiang He, Xinwen Hou, Yu Liu, POPO: Pessimistic Offline Policy Optimization, Offline Reinforcement Learning Workshop on NIPS, 2020.

[37]  Qiang He, Xinwen Hou, Reducing Estimation Bias via Weighted Delayed Deep Deterministic Policy Gradient, The annual IEEE International Conference on Tools with Artificial Intelligence (ICTAI), 2020.

[36]  Chen Gong, Yunpeng Bai, Xinwen Hou, and Xiaohui Ji, Stable Training of Bellman Error in Reinforcement Learning, International Conference on Neural Information Processing(ICONIP), 2020.

[35]  Yekun Chai, Jin Shuo, Xinwen Hou,  Highway Transformer: Self-Gating Enhanced Self-Attentive Networks, Annual Meeting of the Association for Computational Linguistics(ACL), 2020.

[34]  Yingying Chen, Xinwen Hou, An Improvement based on Wasserstein GAN for Alleviating Mode Collapsing, International Joint Conference on Neural Networks(IJCNN), 2020.

[33]  Xiangxiang Shen, Chuanhuan Yin, Yekun Chai and Xinwen Hou, Exponential Moving Averaged Q-network for DDPG, The Second China pattern recognition and computer vision Conference (PRCV), 2019.

[32]  Zhunan Li and Xinwen Hou,Mixing Update Q-value for Deep Reinforcement Learning, International Joint Conference on Neural Networks(IJCNN), 2019.

[31]  Xiangxiang Shen, Chuanhuan Yin, Xinwen Hou, Self-Attention for Deep Reinforcement Leraning, International Conference on Mathematics and Artificial Intelligence(ICMAI), 2019.

[30]  Y. Wang, X.-Y. Zhang, Y. Zhang, X. W. Hou and C.-L Liu, Exploiting Coarse-to-Fine Mechanism for Fine-Grained Recognition, Proc. Int. Conf. Image Processing (ICIP), Phoenix, USA, September 25-28, 2016, pp.649-653.

[29]  D. Wang, X. W. Hou, C.-L Liu, Traffic Sign Detection from Video: A Fast Approach with Tracking, Proc. 3rd ACPR, Kuala Lumpur, Malaysia, 2015.

[28]  D. Wang, S. Yue, J. Xu, X. W. Hou, C.-L Liu, A saliency-based cascade method for fast traffic sign detection, IEEE Intelligent Vehicles Symposium, Seoul, Korea, 2015, pp.180-185.

[27] Y. Liu, X. W. Hou, C.-L. Liu, A compact spatial feature representation for image classification, Proc. 2nd ACPR, Okinawa, Japan, 2013, pp.601-605.

[26] X.-J. Jin, Q.-F. Wang, X. W. Hou, C.-L. Liu, Visual gesture character string recognition by classification-based segmentation with stroke deletion, Proc. 2nd ACPR, Okinawa, Japan, 2013, pp.120-124.

[25] Y. Liu, X.-Y. Zhang, K. Huang, X. W. Hou, C.-L. Liu, Multiple outlooks learning with support vector machines, Proc. ICONIP 2012, LNCS Vol.7665, pp.116-124.

[24]  A. Lu, X. W. Hou, C.-L. Liu, X. Chen, Insect species recognition using discriminative local soft coding, Proc. 21th ICPR, Tsukuba, Japan, 2012, pp.1221-1224.

[23]  Y.-F. Pan, C.-L. Liu, X. W. Hou, Fast scene text localization by learning-based filtering and verification, Proc. Int. Conf. Image Processing (ICIP), Hong Kong, 2010, pp.2269-2272.

[22]  A. Lu, X. W. Hou, X. Chen, C.-L. Liu, Insect species recognition using sparse representation, Proc. BMVC 2010, Aberystwyth, UK, 2010.

[21]  X.-B. Jin, X. W. Hou, C.-L. Liu, Multi-class AdaBoost with hypothesis margin, Proc. 20th ICPR, Istanbul, Turkey, 2010, pp.65-68.

[20]  H. Wang, X. W. Hou, C.-L. Liu, Boosting incremental semi-supervised discriminant analysis for tracking, Proc. 20th ICPR, Istanbul, Turkey, 2010, pp.2748-2751.

[19]  Y.-M. Zhang, Y. Zhang, D.-Y. Yeung, C.-L. Liu, X. W. Hou, Transductive learning on adaptive graphs, Proc. 24th AAAI Conference on Artificial Intelligence, 2010, pp.661-666.

[18]  G. Zhong, W.-J. Li, D.-Y. Yeung, X. W. Hou, C.-L. Liu, Gaussian process latent random field, Proc. 24th AAAI Conference on Artificial Intelligence, 2010, pp.679-684.

[17]  G. Zhong, X. W. Hou, C.-L. Liu, Relative distance-based Laplacian eigenmaps, Proc. 2009 Chinese Conference on Pattern Recognition (CCPR) and First CJK Workshop on Pattern Recognition (CJKPR), Nanjing, China, 2009, pp.784-788.

[16]  H. Wang, X. W. Hou, C.-L. Liu, Object tracking by bidirectional learning with feature selectionICIP2009, Cairo, Egypt, 2009.

[15]  Y.-M. Zhang, X. W. Hou, S. Xiang, C.-L. Liu, Subspace regularization: a new semi-supervised learning methodECML PKDD2009Bled, Slovenia, LNCS Vol.5782, Springer, 2009, pp.586-601.

[14]  Y.-F. Pan, X. W. Hou, C.-L. Liu, Text localization in natural scene images based on conditional random field, Proc. 10th ICDAR, Barcelona, Spain, 2009, pp.6-10.

[13]   X.-B. Jin, C.-L. Liu, X. W. Hou, Prototype learning by margin-based conditional log-likelihood loss, Proc. 19th ICPR, Tampa, USA, 2008.

[12]  Y.-F. Pan, X. W. Hou, C.-L. Liu, A robust system to detect and localize texts in natural scene images, Proc. 8th IAPR Int. Workshop on DAS, Nara, Japan, Sep. 2008, pp.35-42.

[11]  X.-H. Liu, C.-L. Liu, X. W. Hou, A pooled subspace mixture density model for pattern classification in high-dimensional spaces, Proc. IJCNN 2008, Hong Kong, pp.2467-2472.

[10]  X.-B. Jin, X. Hou, C.-L. Liu, A hybrid generative-discriminative learning algorithm for Bayesian network structure, Proc. 5th Int. Conf. on Wavelet Analysis and Pattern Recognition, Beijing, China, 2007, Vol.2, pp.618-623.

[9]  Fei Wu, Yong-Ge Wang and Xin-Wen Hou, License Plate Character Recognition Based on Framelet, Proc. 5th Int. Conf. on Wavelet Analysis and Pattern Recognition, Beijing, China, 2007, Vol.2, pp.673-676.

[8]  X. Hou, C.-L. Liu, Tieniu Tan, Learning boosted asymmetric classifiers for object detection, Proc. CVPR 2006, New York, Vol.1, pp.330-338.

[7]  Y. Wang, X. W. Hou, T. N. Tan, Recognize Multi-people Interactive Activity by PCA-HMMs, the 7th Asian Conference on Computer Vision (ACCV06), Lecture Notes in Computer Science, 2006, vol. 3581, pp. 160-169.

[6]  J. L. Cui, T. N. Tan, X. W. Hou, Y. H. Wang, Z. S. Wei, An Iris Detection Method Based on Structure Information, Advances in Biometric Person Authentication: International Workshop on Biometric Recognition Systems, Lecture Notes in Computer Science, 2005, vol. 3781, pp. 157-164

[5]  Stan Z. Li, Xianhuan Peng, Xinwen Hou, Hongjiang Zhang and Qiansheng Cheng, Multi-view face pose estimation based on supervised ISA learning, Proceedings of th Fifth IEEE International Conference on Automatic Face and Gesture Recognition, 2002.

[4]  Xinwen Hou, Stan Z. Li, Hongjiang Zhang, and Qiansheng Cheng, Direct Appearance Models, IEEE International Conference on Computer Vision and Pattern Recognition, pp.828-833, Vol 1, 2001.

[3]  S. Z. Li, X. W. Hou, H. J. Zhang, Q. S.  Cheng, Learning Spatially Localized, Parts-Based Representation,  Proceedings of IEEE International Conference on Computer Vision and Pattern Recognition, 2001, vol. 1, pp. 207-212.

[2]  S.Z. Li, J. Yan, X.W. Hou, Z.Y. Li, and H.J. Zhang, Learning Low Dimensional Invariant Signature of 3-D Object under Varying View and Illumination from 2-D Appearances, In Proceedings of 8th IEEE International Conference on Computer Vision, Vancouver, Vol.1, pp.635–640, 2001.

[1] Weidong Sun, and Xinwen Hou, Multi-Sensor Information Synthesized Method for Cloud Clearing, Proceedings of 1997 China-Japan Symposium on Advanced Information Technology, 1997, pp. 362-367.


科研活动

   
科研项目
( 1 ) 自底向上的人机对抗技术, 负责人, 研究所自主部署, 2017-01--2019-12
( 2 ) 人机对抗空间建模与态势认知, 参与, 国家任务, 2016-11--2019-10
( 3 ) 2018年人机对抗挑战赛, 参与, 国家任务, 2018-03--2020-03
( 4 ) 2018年人机对抗挑战赛-先知兵圣AI接口与框架开发, 负责人, 国家任务, 2018-03--2020-03
( 5 ) 2019年人机对抗挑战赛, 参与, 国家任务, 2019-06--2021-05
( 6 ) 2019年人机对抗挑战赛-先知兵圣AI接口与框架开发, 负责人, 国家任务, 2019-06--2021-05
( 7 ) 中科院先导专项:任务生成及智能体基准能力测试, 参与, 国家任务, 2020-07--2024-12
( 8 ) 智能试验评估系统有效性验证, 负责人, 国家任务, 2022-09--2023-08
( 9 ) 博弈对抗演练场平台, 负责人, 国家任务, 2022-12--2025-11