Yunkang CAO

Yunkang CAO

Hunan University

Yunkang Cao (曹云康)

I am currently an Assistant Professor at the School of Artificial Intelligence and Robotics, Hunan University (HNU). I am proud to be a core member of the distinguished research team led by Yaonan Wang (王耀南院士) and Hui Zhang (张辉院长).

I received my Ph.D. from Huazhong University of Science and Technology (HUST), mentored by Prof. Weiming Shen (FCAE, FIEEE). During my doctoral studies, I was a visiting researcher at Politecnico di Milano (Polimi) under the supervision of Prof. Giacomo Boracchi.

[Research Vision] My current research passion lies in constructing a deployable, full-chain toolchain for industrial anomaly detection, spanning the entire lifecycle from Perception (Detection) to Cognition (Understanding/Reasoning), and finally to Action (Recovery). I am actively extending these capabilities to the domain of Embodied AI and Robotics, with a specific focus on Unmanned Autonomous Inspection Systems. My ultimate goal is to bridge the gap between academic SOTA models and reliable, open-world industrial applications.

Currently, I serve as the Executive Guest Editor for the SI of Foundation Models for Anomaly Detection, Reasoning, and Recovery in Pattern Recognition.

🌟 Join Us

Let’s Define the Future of Industrial AI

We are actively recruiting Master’s students, Research Assistants (RA), and Visiting Scholars. Highly motivated undergraduates with a strong mathematical background are also welcome to join as interns.

We value genuine problem-solving over metric-chasing. We are looking for partners who have a keen sense for cutting-edge technologies (e.g., AIGC, Foundation Models) and solid coding skills to explore the boundaries of AI in industry and robotics together.

What you can expect:

  • Deep Collaboration: I am not just a supervisor, but a comrade-in-arms on your research journey. We will leverage AIGC and MLLMs to reconstruct traditional industrial vision, tackling system-level challenges from “Perception” to “Recovery”.
  • Diversified Growth: Whether your goal is publishing in top-tier venues (CVPR/AAAI/TPAMI) or solving critical bottlenecks in real-world deployment, I will provide customized guidance to help you achieve impactful results.

🔬 Research Focus

If you are eager to make systematic contributions at the intersection of Industrial Foundation Models and Embodied AI, join us to explore:

  1. Industrial Foundation Models: Investigating the application of Scaling Laws in industrial vision to build efficient, general-purpose backbones. Leveraging the generalization power of MLLMs to solve the Cold Start problem, achieving high-precision detection of both known defects and unknown anomalies simultaneously.
  2. AIGC & Controllable Anomaly Synthesis: Addressing the inherent scarcity of anomaly data by using Generative AI. The goal is to synthesize visually realistic, controllable, and physics-compliant defect samples, boosting model training through high-quality synthetic data.
  3. Multimodal Reasoning & Diagnosis: Moving beyond “Perception” to “Cognition”. Utilizing MLLMs to describe anomalies and analyze their underlying mechanisms. Combining Agentic AI to realize a closed loop from problem discovery to autonomous decision-making and repair.
  4. Embodied AI & Autonomous Inspection: Deploying visual foundation models onto unmanned systems (e.g., mobile robots, manipulators) to enable autonomous perception in open-world environments. Focusing on identifying generalized anomalies that violate “safety states” in unstructured scenarios.

📩 Contact: Please send your CV to caoyunkang0207@gmail.com

🌟 加入我们

定义工业 AI 的未来：从基础模型到具身智能

本课题组长期招收硕士研究生、科研助理（RA）及访问学者，亦欢迎数理基础扎实的本科生提前进组实习。

我们坚持以解决真问题为核心，相比于单纯的“刷榜”，我们更看重技术背后的逻辑与实际价值。如果你对 AIGC、Foundation Models 等前沿技术充满好奇，且拥有扎实的代码落地能力，欢迎加入我们，共同拓展 AI 在工业与机器人领域的边界。

在这里，你将获得：

  • 并肩作战 在这里，我会成为你科研路上的战友。我将带你深入一线，利用 AIGC 和 MLLMs 重构传统的工业视觉范式，共同挑战从“感知诊断”到“自主修复”的系统级难题。

  • 个性化成长 无论你志在冲击顶会（CVPR/AAAI/TPAMI），还是渴望解决工业界的“卡脖子”难题，我都会结合你的特质定制培养方案，让你的成果既有学术高度，又有落地回响。

🔬 重点研究方向

如果你渴望在以下工业大模型与具身智能的交叉领域做出系统性贡献，欢迎加入我们：

  1. 工业视觉基础模型 探索 Scaling Law 在工业视觉领域的应用，构建高效、通用的工业视觉底座。致力于利用 MLLMs (多模态大语言模型) 强大的泛化能力，解决工业场景下的冷启动 (Cold Start) 难题，实现单一大模型对“已知缺陷”与“未知异常”的同时高精度检测。
  2. AIGC 与可控异常生成 针对工业异常数据天然稀缺的痛点，研究基于 AIGC 的数据增强技术。目标是合成视觉真实、语义可控且符合物理规律 (Physics-compliant) 的缺陷样本，通过高质量的合成数据辅助模型训练，突破数据瓶颈。
  3. 多模态推理与智能诊断 超越传统的“感知”边界，向“认知”延伸。利用 MLLMs 对异常进行深层语义描述与成因机理分析。结合 智能体 (Agents) 技术，实现从“发现问题”到“自主决策”再到“闭环修复”的完整工业智能链路。
  4. 具身智能与自主巡检 将视觉大模型部署于无人系统（如移动机器人、机械臂），赋予机器人在开放环境下的自主感知与交互能力。重点研究非结构化场景中对不符合“安全状态”的广义异常识别，实现真正的具身智能巡检。

📩 联系方式: 请发送简历至 caoyunkang0207@gmail.com

🔥 News

📝 Selected Publications

(For the complete list of publications, please refer to My Google Scholar Page)

📝 Selected Publications

Citations Top Tier Papers GitHub Stars

(For the complete list of publications, please refer to My Google Scholar Page) # co-first author | * corresponding author

  1. Towards zero-shot point cloud anomaly detection: A multi-view projection framework [Paper] [Code]
    Yuqi Cheng#, Yunkang Cao#, Guoyang Xie, Zhichao Lu, Weiming Shen*
    IEEE Transactions on Systems, Man, and Cybernetics: Systems (IEEE TSMC). 2026.
  2. A Comprehensive Survey for Real-World Industrial Defect Detection: Challenges, Approaches, and Prospects [Paper] [Code]
    Yuqi Cheng#, Yunkang Cao#, Haiming Yao, Wei Luo, Cheng Jiang, Hui Zhang, Weiming Shen*
    Journal of Manufacturing Systems (JMS). 2026. (In Press)
  3. IAD-R1: Reinforcing Consistent Reasoning in Industrial Anomaly Detection
    Yanhui Li, Yunkang Cao, Chengliang Liu, Yuan Xiong, Xinghui Dong, Chao Huang [Paper] [Code]
    The 40th Annual AAAI Conference on Artificial Intelligence (AAAI). 2026. (Oral Presentation)
  4. Towards High-resolution 3D Anomaly Detection: A Scalable Dataset and Real-time Framework for Subtle Industrial Defects [Paper] [Code]
    Yuqi Cheng, Yihan Sun, Hui Zhang, Weiming Shen, Yunkang Cao*
    The 40th Annual AAAI Conference on Artificial Intelligence (AAAI). 2026. (Oral Presentation)
  5. Anomagic: Cross-modal Prompt-driven Zero-shot Anomaly Generation [Paper] [Code]
    Yuxin Jiang, Wei Luo, Hui Zhang, Qiyu Chen, Haiming Yao, Weiming Shen, Yunkang Cao*
    The 40th Annual AAAI Conference on Artificial Intelligence (AAAI). 2026. (Poster Presentation)
  6. Global-Regularized Neighborhood Regression for Efficient Zero-Shot Texture Anomaly Detection [Paper] [Code]
    Haiming Yao, Wei Luo, Yunkang Cao, Yiheng Zhang, Wenyong Yu*, Weiming Shen
    IEEE Transactions on Systems, Man, and Cybernetics: Systems (IEEE TSMC).
  7. Exploring Intrinsic Normal Prototypes within a Single Image for Universal Anomaly Detection [Paper] [Code]
    Wei Luo#, Yunkang Cao#, Haiming Yao#, Xiaotian Zhang, Jianan Lou, Yuqi Cheng, Weiming Shen, Wenyong Yu*
    IEEE Conference on Computer Vision and Pattern Recognition (CVPR). 2025.
  8. Anomaly Anything: Promptable Unseen Visual Anomaly Generation [Paper] [Code]
    Han Sun, Yunkang Cao, Hao Dong, Olga Fink*
    IEEE Conference on Computer Vision and Pattern Recognition (CVPR). 2025.
  9. Customizing Visual-Language Foundation Models for Multi-Modal Anomaly Detection and Reasoning [Paper] [Code]
    Xiaohao Xu#, Yunkang Cao#, Huaxin Zhang, Nong Sang, Xiaonan Huang, Weiming Shen*
    IEEE International Conference on Computer Supported Cooperative Work in Design (CSCWD). 2025. Best Student Paper Award
  10. Personalizing Vision-Language Models with Hybrid Prompts for Zero-Shot Anomaly Detection [Paper] [Code]
    Yunkang Cao, Xiaohao Xu, Yuqi Cheng, Chen Sun, Zongwei Du, Liang Gao, Weiming Shen*
    IEEE Transactions on Cybernetics (IEEE TCYB). 2025.
  11. VarAD: Lightweight High-Resolution Image Anomaly Detection via Visual Autoregressive Modeling [Paper] [Code]
    Yunkang Cao, Haiming Yao, Wei Luo, Weiming Shen*
    IEEE Transactions on Industrial Informatics (IEEE TII). 2025.
  12. AdaCLIP: Adapting CLIP with Hybrid Learnable Prompts for Zero-Shot Anomaly Detection [Paper] [Code]
    Yunkang Cao, Jiangning Zhang, Luca Frittoli, Yuqi Cheng, Weiming Shen*, Giacomo Boracchi
    European Conference on Computer Vision (ECCV). 2024.
  13. Complementary pseudo multimodal feature for point cloud anomaly detection [Paper] [Code]
    Yunkang Cao, Xiaohao Xu, Weiming Shen*
    Pattern Recognition (PR). 2024.
  14. BiaS: Incorporating Biased Knowledge to Boost Unsupervised Image Anomaly Localization [Paper] [Code]
    Yunkang Cao, Xiaohao Xu, Chen Sun, Liang Gao, Weiming Shen*
    IEEE Transactions on Systems, Man, and Cybernetics: Systems (IEEE TSMC). 2024.
  15. Collaborative discrepancy optimization for reliable image anomaly localization [Paper] [Code]
    Yunkang Cao, Xiaohao Xu, Zhaoge Liu, Weiming Shen*
    IEEE Transactions on Industrial Informatics (IEEE TII). 2023.
  16. High-Resolution Image Anomaly Detection via Spatiotemporal Consistency Incorporated Knowledge Distillation [Paper]
    Yunkang Cao, Yiheng Zhang, Weiming Shen*
    IEEE International Conference on Automation Science and Engineering (IEEE CASE). 2023.
  17. Informative knowledge distillation for image anomaly segmentation [Paper] [Code]
    Yunkang Cao, Qian Wan, Weiming Shen*, Liang Gao
    Knowledge-Based Systems (KBS). 2022.
  18. Boosting Global-Local Feature Matching via Anomaly Synthesis for Multi-Class Point Cloud Anomaly Detection
    Yuqi Cheng, Yunkang Cao, Dongfang Wang, Weiming Shen*, Wenlong Li
    IEEE Transactions on Automation Science and Engineering (IEEE TASE). 2025.
  19. Prototypical Learning Guided Context-Aware Segmentation Network for Few-Shot Anomaly Detection [Paper] [Code]
    Yuxin Jiang, Yunkang Cao, Weiming Shen*
    IEEE Transactions on Neural Networks and Learning Systems (IEEE TNNLS). 2024.
  20. LogiCode: an LLM-Driven Framework for Logical Anomaly Detection [Paper] [Code]
    Yiheng Zhang, Yunkang Cao, Xiaohao Xu, Weiming Shen*
    IEEE Transactions on Automation Science and Engineering (IEEE TASE). 2024.
  21. Prior Normality Prompt Transformer for Multi-class Industrial Image Anomaly Detection [Paper]
    Haiming Yao, Yunkang Cao, Wei Luo, Weihang Zhang, Wenyong Yu*, Weiming Shen
    IEEE Transactions on Industrial Informatics (IEEE TII). 2024.
  22. Deep Feature Contrasting for Industrial Image Anomaly Segmentatio [Paper]
    Qian Wan, Yunkang Cao, Liang Gao, Xinyu Li*, Yiping Gao
    IEEE Transactions on Instrumentation and Measurement (IEEE TIM). 2024.
  23. Dual-path Frequency Discriminators for Few-shot Anomaly Detection
    Yuhu Bai#, Jiangning Zhang#, Zhaofeng Chen, Yuhang Dong, Yunkang Cao, Guanzhong Tian*
    Knowledge-Based Systems (KBS). 2024.
  24. Generative Denoise Distillation: Simple Stochastic Noises Induce Efficient Knowledge Transfer for Dense Prediction
    Zhaoge Liu, Xiaohao Xu, Yunkang Cao, Weiming Shen*
    Knowledge-Based Systems (KBS). 2024.
  25. RAD: A Comprehensive Dataset for Benchmarking the Robustness of Image Anomaly Detection [Paper] [Code]
    Yuqi Cheng, Yunkang Cao, Rui Chen, Weiming Shen*
    IEEE International Conference on Automation Science and Engineering (IEEE CASE). 2024.
  26. Attention Fusion Reverse Distillation for Multi-Lighting Image Anomaly Detection [Paper]
    Yiheng Zhang, Yunkang Cao, Tianhang Zhang, Weiming Shen*
    IEEE International Conference on Automation Science and Engineering (IEEE CASE). 2024.
  27. A masked reverse knowledge distillation method incorporating global and local information for image anomaly detection [Paper] [Code]
    Yuxin Jiang, Yunkang Cao, Weiming Shen*
    Knowledge-Based Systems (KBS). 2023.
  28. Position encoding enhanced feature mapping for image anomaly detection [Paper] [Code]
    Qian Wan, Yunkang Cao, Liang Gao, Weiming Shen, Xinyu Li*
    IEEE International Conference on Automation Science and Engineering (IEEE CASE). 2022.

🥇 Selected Awards

  • National Scholarship (the highest scholarship for Ph.D.), 2024.11
  • Provincial Second Prize, China International College Students’ Innovation Competition, 2024.08
  • 2nd place in CVPR VAND Zero-shot Anomaly Detection Challenge
  • First-class Scholarship for Postgraduates, HUST, 2020.09, 2021.09, 2022.09
  • Student Award for Research and Innovation, HUST, 2022.05
  • Mathematical Modeling Stars Nomination (TOP 2) in the 18th China Post-graduate Mathematical Contest, 2022.05
  • Merit Postgraduate student, HUST, 2021.09
  • Excellent Graduates, HUST, 2019.06
  • National Scholarship (the highest scholarship for B.E), 2017.09, 2019.09
  • Second Class Prize, Undergraduate Electronics Design Contest, Provincial, 2018.09
  • Third Class Prize, Undergraduate Intelligent Robotics Contest, National, 2018.05

🎓📚 Academic Service

  • Executive Guest Editor, Pattern Recognition (Elsevier), Special Issue on “Foundation Models for Anomaly Detection, Reasoning, and Recovery”.
  • Journal Reviewer, IEEE TSMC, IEEE TNNLS, IEEE TII, IEEE TKDE, IEEE TCSVT, IEEE TASE, PR, etc.
  • Conference Reviewer, CVPR, ICCV, ECCV, AAAI, NeurIPS, ICLR, ICRA, IROS.
  • Co-organizer of special sessions, [Anomaly Detection with Foundation Models (ADFM)] at IJCAI (2024).
  • Co-organizer of special sessions, [Industrial Foundation Models and Applications in Smart Manufacturing] at the IEEE International Conference on Automation Science and Engineering (2024).

💬 Invited Talks

  • 2025.11.23, Chongqing University (CQU), “Towards General Visual Anomaly Detection” (Invited by Prof. Yan Qin).
  • 2025.11.07, Central South University (CSU), “Towards General Visual Anomaly Detection” (Invited by Prof. Senzhang Wang).
  • 2025.10.30, Sun Yat-sen University (SYSU) & Tencent Youtu Lab, “Towards General Visual Anomaly Detection” (Invited by Prof. Chao Huang).
  • 2025.10.12, Xiangtan, “Towards General Visual Anomaly Detection”.
  • 2025.07.28, Lanzhou, “Towards General Visual Anomaly Detection”.
  • 2024.07, EPFL, “Application-Oriented Industrial Visual Anomaly Detection” [Slides].
  • 2023.11, National University of Defense Technology (NUDT), “Overview of Image Anomaly Detection—Review, Applications, and Future Prospects” [Slides].

📖 Education

  • 2023.10 - 2024.10, Politecnico di Milano

    Department of Electronics, Information and Bioengineering
    Visiting Ph.D. in Computer Science                                      Advisor: Giacomo Boracchi

  • 2020.09 - 2025.06, Huazhong University of Science and Technology

    State Key Laboratory of Digital Manufacturing Equipment and Technology
    Ph.D. Candidate in Mechanical Engineering                                      Advisor: Weiming Shen

  • 2016.09 - 2020.06, Huazhong University of Science and Technology

    B.S. in Mechanical Design, Manufacture & Automation    

📋 Work Experience

  • 2025.06 - Present, Hunan University

    Assistant Professor, School of Artificial Intelligence and Robotics, Hunan University