PhD Jinbao Wang

Assistant Professor, National Engineering Laboratory for Big Data System Computing Technology, Shenzhen University (SZU). I work on digital human modelling and driving, image anomaly detection, computer vision and machine learning.

CV  /  CV(中文)  /  Bio  /  Github  /  Google Scholar   

If you have any questions or wish to cooperate with us, please feel free to contact me. wangjb@szu.edu.cn / linkingring@163.com / wangjb@ieee.org

I'm interested in image anomaly detection, graph representation learning, computer vision, and machine learning. Much of my research is about detecting and localizing anomalies from images. Representative papers are highlighted.

Note that *contributed equally, †corresponding author

🔥       Image Anomaly Detection

We provide a comprehensive review of deep learning-based IAD from the perspectives of neural network architectures, levels of supervision, loss functions, metrics and datasets.

We propose a large-scale systematic benchmark and uniform setting for IAD to bridge the gap between academy and industrial manufacturing.

To achieve better representation, we pose a key question: how to create an ideal distribution required by HRPCD-AD in the feature space? This paper propose a novel group-level feature-based network, called Group3AD, which has a significantly efficient representation ability. The experimental result surpasses Reg3D-AD by the margin of 5% in terms of object-level AUROC on Real3D-AD.

This project introduces a novel Unsupervised Continual Anomaly Detection framework (UCAD), which equips the unsupervised AD with continual learning capability through contrastively-learned prompts.

This project aims to construct a new dataset of high-resolution 3D point clouds for anomaly detection tasks in real-world scenes. Real3D-AD comprises a total of 1,254 samples that are distributed across 12 distinct categories.

This paper addresses a promising and challenging task, i.e., deployment-friendly 3D-AD and proposes an easy but effective neural network (EasyNet) to achieve competitive performance without using large pre-trained models and memory banks.

We reveal that rotation-invariant feature property has a significant impact in industrial-based fewshot anomaly detection.

We systematically investigate various data augmentation methods for few-shot IAD algorithms.

We propose a memory-based unsupervised AD method, which efficiently denoises the data at the patch level.

We propose a unified framework by incorporating continual learning to achieve our newly designed task of continual anomaly detection.

🤹       Digital Human

Hair editing is a critical image synthesis task that aims to edit hair color and hairstyle using text descriptions or reference images, while preserving irrelevant attributes (e.g., identity, background, cloth). This paper utilizes Latent Diffusion Models (LDMs) for hairstyle editing. The proposed method not only tackles the complexity of multi-color hairstyles but also addresses the challenge of preserving original colors during diffusion editing.

The recent methods for deep 3D pose estimation are categorized and thoroughly analyzed. Provide an extensive review of related datasets and evaluation metrics. Compare the pros and cons of the deep 3D models valuated on the datasets and draw a conclusion.

A novel application system is proposed to achieve the generation of 3D character animation driven by markerless human body motion capture.

👑       Medical Federated Learning

We provide a comprehensive review of cross-modality synthesis for neuroimages, from the perspectives of weakly-supervised and unsupervised settings, loss functions, evaluation metrics, ranges of modality, datasets, and the synthesis-based downstream applications.

We propose a new benchmark for federated domain translation on unsupervised brain image synthesis to bridge the gap between federated learning and medical GAN.

We propose a method that can reduce the demands for deformable registration while encourage to leverage the misaligned and unpaired data.

🐰       Fast Retrieval

This paper proposes a novel yet effective framework coined cross-modal Alternating Learning with Task-Aware representations (ALTA) to make good use of visual and linguistic modal information and achieve more effective continual learning.

We propose a novel framework for the new task of continuous cross-modal hashing.

We propose a multi-objective oneshot network architecture search algorithm to obtain the best trade-off networks in terms of the adversarial accuracy, the clean accuracy and the model size.

To represent graph-based data, and maintain fast retrieval while doing so, we introduce an efficient hash model with graph neural networks for fast graph-based data retrieval.

☘️       Image Propcessing

We propose seminar learning, a new learning paradigm for semantic segmentation with click-level supervision.

We propose a single image dehazing method based on a physical model and the brightness components of the image by using a multi-scale retinex with color restoration algorithm.

We propose a single image dehazing method that is based on a physical model and the dark channel prior principle.

Note that * contributed equally, † corresponding authors.

Conference

1. Yu Zeng, Yang Zhang, Linlin Shen, Jiachen Liu, Kaijun Deng, Weizhao He, and Jinbao Wang. "HairDiffusion: Vivid Multi-Colored Hair Editing via Latent Diffusion." The Thirty-Eighth Annual Conference on Neural Information Processing Systems (NeurIPS). 2024.   (CCF-A)  
2. Teng Yang, Pengcheng Gao, Chengbin Hou, Jinbao Wang, and Yongliang Tang. "Self-supervised Models are Strong Industrial Few-shot Defect Classification Learners." 2nd workshop on Vision-based InduStrial InspectiON. ECCVW-VISION. 2024.
3. Qingyuan Liu, Ke Lv, Zehai Niu, Kun Dong, Jinbao Wang, Jian Xue, and Xiaoyu Qin. "FlexControl: Flexible and Efficient Full-Body Controllable Text-to-Motion Generation." Towards a Complete Analysis of People: Fine-grained Understanding for Real-World Applications. ECCVW-TCAP. 2024.
4. Qingyuan Liu, Zehai Niu, Ke Lu, Kun Dong, Jian Xue, Xiaoyu Qin, and Jinbao Wang. "AdaptControl: Adaptive Human Motion Control and Generation via User Prompt and Spatial Trajectory Guidance." The 5th International Workshop on Human-centric Multimedia Analysis. ACM MMW-HCMA. 2024.   (Best Student Paper)  
5. Hongze Zhu, Guoyang Xie, Chengbin Hou, Tao Dai, Can Gao, Jinbao Wang†, and Linlin Shen. "Towards High-resolution 3D Anomaly Detection via Group-Level Feature Contrastive Learning." In Proceedings of the 32st ACM International Conference on Multimedia (ACM MM). 2024.   (CCF-A)  
6. Xinpeng Li, Teng Wang, Shuyi Mao, Jinbao Wang, Jian Zhao, Xiaojiang Peng, Feng Zheng, and Xuelong Li. "Two in One Go: Single-stage Emotion Recognition with Decoupled Subject-context Transformer." In Proceedings of the 32st ACM International Conference on Multimedia (ACM MM). 2024.   (CCF-A)  
7. Lian Chen, Zehai Niu, Qingyuan Liu, Jinbao Wang, Jian Xue, and Ke Lu. "Anatomically-informed vector quantization variational auto-encoder for text to motion generation." IEEE International Conference on Multimedia and Expo, Workshop (ICMEW). 2024.
8. Tao Dai, Jianping Wang, Hang Guo, Jinmin Li, Jinbao Wang†, and Zexuan Zhu†. "FreqFormer: Frequency-aware Transformer for Lightweight Image Super-resolution." International Joint Conference on Artificial Intelligence (IJCAI). 2024.   (CCF-A)  
9. Qiang Li, Qianchen Mao, Wenjie Liu, Jinbao Wang, Wenming Wang, and Binshu Wang. "Local Information Guided Global Integration For Infrared Small Target Detection." IEEE International Conference on Acoustics, Speech, and Signal Processing (ICASSP). 2024.
10. Jiaqi Liu, Kai Wu, Qiang Nie, Ying Chen, Bin-Bin Gao, Yong Liu, Jinbao Wang, Chengjie Wang, and Feng Zheng. "Unsupervised Continual Anomaly Detection with Contrastively-learned Prompt." Association for the Advancement of Artificial Intelligence (AAAI). 2024.
11. Jiaqi Liu, Guoyang Xie, Ruitao Chen, Xinpeng Li, Jinbao Wang†, Yong Liu, Chengjie Wang, and Feng Zheng†. "Real3D-AD: A Dataset of Point Cloud Anomaly Detection." Datasets & Benchmarks Track, Neural Information Processing Systems (NeurIPS). 2023.   (CCF-A)  
12. Ruitao Chen, Guoyang Xie, Jiaqi Liu, Jinbao Wang†, Ziqi Luo, Jinfan Wang, Feng Zheng†. "EasyNet: An Easy Network for 3D Industrial Anomaly Detection." In Proceedings of the 31st ACM International Conference on Multimedia (ACM MM). 2023.   (CCF-A)  
13. Lingrui Zhang, Shuheng Zhang, Guoyang Xie, Jiaqi Liu, Hua Yan, Jinbao Wang†, Feng Zheng†, and Yaochu Jin. "What makes a good data augmentation for few-shot unsupervised image anomaly detection?" In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR Vision Workshop), pp. 4344-4353. 2023.
14. Guoyang Xie*, Jinbao Wang*†, Jiaqi Liu*, Feng Zheng†, and Yaochu Jin. "Pushing the limits of fewshot anomaly detection in industry vision: Graphcore." The Eleventh International Conference on Learning Representations (ICLR). 2023.
15. Wujin Li, Jiawei Zhan, Jinbao Wang†, Bizhong Xia, Bin-Bin Gao, Jun Liu, Chengjie Wang, and Feng Zheng†. "Towards continual adaptation in industrial anomaly detection." In Proceedings of the 30th ACM International Conference on Multimedia (ACM MM), pp. 2871-2880. 2022.   (CCF-A)  
16. Xi Jiang, Jianlin Liu, Jinbao Wang†, Qiang Nie, Kai Wu, Yong Liu, Chengjie Wang, and Feng Zheng†. "SoftPatch: Unsupervised anomaly detection with noisy data." Advances in Neural Information Processing Systems (NeurIPS), 35, pp. 15433-15445. 2022.   (CCF-A)  
17. Jinbao Wang*, Guoyang Xie*, Yawen Huang*, Yefeng Zheng, Yaochu Jin, and Feng Zheng. "FedMed-ATL: Misaligned unpaired cross-modality neuroimage synthesis via affine transform loss." In Proceedings of the 30th ACM International Conference on Multimedia (ACM MM), pp. 1522-1531. 2022.   (CCF-A)  
18. Hongjun Chen, Jinbao Wang, Hong Cai Chen, Xiantong Zhen, Feng Zheng, Rongrong Ji, and Ling Shao. "Seminar learning for click-level weakly supervised semantic segmentation." In Proceedings of the IEEE/CVF International Conference on Computer Vision (ICCV), pp. 6920-6929. 2021.   (CCF-A)  
19. Lian Chen, Ke Lu, Pengcheng Gao, Jian Xue, and Jinbao Wang. "A novel multi-feature skeleton representation for 3d action recognition." In International Conference on Pattern Recognition (ICPR), pp. 365-379. Springer, Cham, 2021.
20. Jinbao Wang, Ke Lu, Jian Xue, and Yutong Kou. "Relative depth estimation prior for single image dehazing." In 2019 IEEE International Conference on Multimedia & Expo Workshops (ICMEW), pp. 270-275. IEEE, 2019.
21. Jinbao Wang, Ke Lu, Jian Xue, Pengcheng Gao, and Yanfu Yan. "A markerless body motion capture system for character animation based on multi-view cameras." In ICASSP 2019-2019 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP), pp. 8558-8562. IEEE, 2019.
22. Jinbao Wang, Ning He, and Ke Lu. "A new single image dehazing method with MSRCR algorithm." In Proceedings of the 7th International Conference on Internet Multimedia Computing and Service, pp. 1-4. 2015.
23. Ning He, Ke Lu, and Jinbao Wang. "Image denoising using fractional-order non-local TV model." In Proceedings of International Conference on Internet Multimedia Computing and Service, pp. 279-282. 2014.

Journal

1. Bingshu Wang, Haosu Zhang, Qiang Li, Qianchen Mao, Jinbao Wang, C.L. Philip Chen, Aihong Shangguan, and Haosu Zhang. "A Survey on Vision-Based Anti Unmanned Aerial Vehicles Methods." Drones. 2024. IF
2. Zehai Niu, Ke Lu, Jian Xue, Xiaoyu Qin, Jinbao Wang, and Ling Shao. "From Method to Application: A Review of Deep 3D Human Motion Capture." IEEE Transactions on Circuits and Systems for Video Technology (IEEE TCSVT). 2024. IF   (中科院大类1区)  
3. Zehai Niu, Ke Lu, Jian Xue, and Jinbao Wang. "Skeleton Cluster Tracking for robust multi-view multi-person 3D human pose estimation." Computer Vision and Image Understanding (CVIU). 2024. IF
4. Guoyang Xie*, Jinbao Wang*, Jiaqi Liu*, Jiayi Lyu, Yong Liu, Chengjie Wang, Feng Zheng, and Yaochu Jin. "IM-IAD: Industrial image anomaly detection benchmark in manufacturing." IEEE Transactions on Cybernetics (IEEE TCYB). 2024. IF   (中科院大类1区)  
5. Jiaqi Liu, Guoyang Xie, Jinbao Wang*, Shangnian Li, Chengjie Wang, Feng Zheng, and Yaochu Jin. "Deep industrial image anomaly detection: A survey." Machine Intelligence Research (MIR) 21, no. 1 (2024): 104-135. IF
6. Wujin Li, Bin-Bin Gao, Bizhong Xia, Jinbao Wang, Jun Liu, Yong Liu, Chengjie Wang, and Feng Zheng. "Cross-Modal Alternating Learning with Task-Aware Representations for Continual Learning." IEEE Transactions on Multimedia (IEEE TMM). 2023. IF   (中科院大类1区)  
7. Guoyang Xie*, Yawen Huang*, Jinbao Wang†, Jiayi Lyu, Feng Zheng†, Yefeng Zheng, and Yaochu Jin. "Cross-modality neuroimage synthesis: A survey." ACM Computing Surveys 56 (2023): 1-28. IF   (中科院大类1区)  
8. Guoyang Xie*, Jinbao Wang*, Guo Yu, Feng Zheng, and Yaochu Jin. "Tiny adversarial mulit-objective oneshot neural architecture search." Complex & Intelligent Systems (CIS) 6 (2023): 107-109. IF
9. Jinbao Wang*, Xie, Guoyang*, Yawen Huang*, Jiayi Lyu, Feng Zheng, Yefeng Zheng, and Yaochu Jin. "FedMed-GAN: Federated domain translation on unsupervised cross-modality brain image synthesis." Neurocomputing 546 (2023): 126282. IF
10. Hao Zheng*, Jinbao Wang*, Xiantong Zhen, Jingkuan Song, Feng Zheng, Ke Lu, and Guo-Jun Qi. "Continuous cross-modal hashing." Pattern Recognition (PR) 142 (2023): 109662. IF   (中科院大类1区)  
11. Jinbao Wang, Shuo Xu, Feng Zheng, Ke Lu, Jingkuan Song, and Ling Shao. "Learning efficient hash codes for fast graph-based data similarity retrieval. "IEEE Transactions on Image Processing (IEEE TIP) 30 (2021): 6321-6334. IF   (CCF-A)     (中科院大类1区)  
12. Jinbao Wang*, Shujie Tan*, Xiantong Zhen, Shuo Xu, Feng Zheng, Zhenyu He, and Ling Shao. "Deep 3D human pose estimation: A review." Computer Vision and Image Understanding (CVIU) 210 (2021): 103225. IF
13. Jinbao Wang, Ke Lu, Jian Xue, Ning He, and Ling Shao. "Single image dehazing based on the physical model and MSRCR algorithm." IEEE Transactions on Circuits and Systems for Video Technology (IEEE TCSVT) 28, no. 9 (2017): 2190-2199. IF   (中科院大类1区)  
14. Jinbao Wang, Ning He, Lulu Zhang, and Ke Lu. "Single image dehazing with a physical model and dark channel prior." Neurocomputing 149 (2015): 718-728. IF   (ESI Highly-Cited Paper)  
15. Ning He, Jinbao Wang, Lu-Lu Zhang, and Ke Lu. "An improved fractional-order differentiation model for image denoising." Signal Processing 112 (2015): 180-188. IF
16. Ning He, Jinbao Wang, Lu-Lu Zhang, Guang-Mei Xu, and Ke Lu. "Non-local sparse regularization model with application to image denoising." Multimedia Tools and Applications 75, no. 5 (2016): 2579-2594. IF
17. Ning He, Jinbao Wang, Lu-Lu Zhang, and Ke Lu. "Convex optimization based low-rank matrix decomposition for image restoration." Neurocomputing 172 (2016): 253-261. IF
18. Ning He, Ke Lu, Bing-Kun Bao, Lu-Lu Zhang, and Jinbao Wang. "Single-image motion deblurring using an adaptive image prior." Information Sciences 281 (2014): 736-749. IF