城市级视频监控系统是平安城市建设的重要内容，已从第一代“看得见”和第二代“看得清”的阶段发展到“看得懂”的第三阶段。由于城市级摄像头网络空间分布广、设备类型和参数差异大，实现城市级视频监控系统“看得懂”面临重大挑战。为此，课题组重点研究图像跨域重建与可信身份认真技术，具体包括（1）行为分析（异常行为检测、行为定位与识别等）：通过多尺度边界敏感的时序行为定位与识别思路提取完整语义信息，通过基于时域-空域融合特征的检测思路来差异化正常与异常数据的重构质量；（2）跨模态行人重识别：对跨模态行人高维特征的模态差异进行度量与约束以提高特征的模态不变性；（3）视频目标清晰化（底层视觉）：联合先验信息与运动不变性以提升帧间时序依赖关系的表征能力；（4）跨域图像合成（异质人脸图像生成、图像风格化等）：将不同模态的图像转化为统一模态的图像，实现信息补全；（5）跨域图像识别（异质人脸图像识别），通过表征解构学习提高跨域图像合成与识别的可解释性和精确度。该课题的研究能为网络视频大数据智能分析提供系统的解决方案；（6）可信身份认证：可信这里主要指可靠与安全，算法应该既能防御来自外部的攻击（对抗学习），又能对隐私信息进行保护。

（一）异质人脸图像合成

代表性论文

1. Mingrui Zhu, Jie Li, Nannan Wang, and Xinbo Gao. Learning Deep Patch Representations for Probabilistic Graphical Model-Based Face Sketch Synthesis.International Journal of Computer Vision(IJCV), 2021.
2. Mingrui Zhu, Jie Li, Nannan Wang, and Xinbo Gao.A Deep Collaborative Framework for Face Photo-Sketch Synthesis.IEEE Transactions on Neural Networks and Learning Systems (IEEE TNNLS), vol. 30, no. 10, pp. 3096-3108, 2019.
3. Mingrui Zhu, Changcheng Liang,Nannan Wang, Xiaoyu Wang, Zhifeng Li, Xinbo Gao. A Sketch-Transformer Network for Face Photo-Sketch Synthesis,International Joint Conference on Artificial Intelligence(IJCAI), 2021.

（二）异质人脸图像识别

代表性论文

1. Chunlei Peng, Xinbo Gao, Nannan Wang, and Jie Li. Graphical Representation for Heterogeneous Face Recognition.IEEE Transactions on Pattern Analysis and MachineIntelligence(IEEE TPAMI), vol. 39, no. 2, pp. 301-312, 2017.
2. Chunlei Peng, Nannan Wang, J Li, X Gao. Re-ranking High-Dimensional Deep Local Representation for NIR-VIS Face Recognition,IEEE Transactions on Image Processing(IEEE TIP), vol. 28, no. 9, pp. 4553-4565, 2019
3. Chunlei Peng, Nannan Wang, Jie Li, Xinbo Gao. Soft Semantic Representation for Cross-Domain Face Recognition,IEEE Transactions on Information Forensics and Security(IEEE TIFS), vol.16, 346-360, 2021.

（三）轻量化神经网络in底层视觉任务

代表性论文

1. Jingwei Xin, Nannan Wang, Xinrui Jiang, Jie Li, Heng Huang, Xinbo Gao. Binarized Neural Network for SingleImage Super-Resolution.European Conference on Computer Vision(ECCV), 2020.
2. Xinrui Jiang, Nannan Wang, Jingwei Xin, Keyu Li, Xi Yang, Xinbo Gao. Training Binary Neural Network without Batch Normalization for Image Super-Resolution,35^thAAAI Conference on Artificial Intelligence(AAAI), 2021.
3. Jingwei Xin, Jie Li, Xinrui Jiang, Nannan Wang, Heng Huang, Xinbo Gao. Wavelet-based Dual Recursive Network for Image Super-Resolution.IEEE Transactions on Neural Networks and Learning Systems(IEEE TNNLS), 2021

（四）跨模态行人重识别

代表性论文

1. Ziyu Wei, Xi Yang, Nannan Wang, Xinbo Gao. Syncretic Modality Collaborative Learning for Visible Infrared Person Re-Identification,International Conference on Computer Vision(ICCV), 2021.
2. Yi Hao, Nannan Wang, Xinbo Gao, Jie Li, and Xiaoyu Wang. Dual-alignment Feature Embedding for Cross-modality Person Re-identification,ACM Multimedia(ACM MM), 2019.
3. Yi Hao, Nannan Wang, Jie Li, Xinbo Gao. HSME: Hypersphere Manifold Embedding Network for Person Re-identification.Thirty-Third AAAI Conference on Artificial Intelligence (AAAI), 2019.

（五）行为分析

代表性论文

1. Xinpeng Ding, Nannan Wang, Shiwei Zhang, De Cheng, Xiaomeng Li, Ziyuan Huang, Mingqian Tang, Xinbo Gao. Support-Set Based Cross-Supervision for Video Grounding,International Conference on Computer Vision(ICCV), 2021.
2. Xinpeng Ding, Nannan Wang, Xinbo Gao, Jie Li, Xiaoyu Li, and Tongliang Liu. KFC: An Efficient Framework for Semi-supervised Temporal Action Localization.IEEE Transactions on Image Processing(IEEE TIP), 2021.

（六）对抗学习

代表性论文

1. Dawei Zhou, Nannan Wang, Chunlei Peng, Xinbo Gao, Xiaoyu Wang, Jun Yu, Tongliang Liu. Removing Adversarial Noise in Class Activation Feature Space,International Conference on Computer Vision(ICCV), 2021.
2. Dawei Zhou, Tongliang Liu, Bo Han, Nannan Wang, Chunlei Peng, Xinbo Gao. Towards Defending Against Adversarial Examples via Attack-Invariant Features,International Conference on Machine Learning(ICML), 2021.

（七）图像风格化

代表性论文

1. Tianwei Lin, Zhuoqi Ma, Fu Li, Dongliang He, Xin Li, Errui Ding, Nannan Wang, Jie Li, Xinbo Gao. Drafting and Revision: Laplacian Pyramid Network for Fast High-Quality Artistic Style Transfer,IEEE Conference on Computer Vision and Pattern Recognition (CVPR),2021.
2. Zhuoqi Ma, Nannan Wang, Jie Li, Xinbo Gao. From Reality to Perception: Genre-based Neural Image Style Transfer.International Joint Conference on Artificial Intelligence (IJCAI), pp.3491-3497, 2018