🌐 欢迎来到我的个人主页！我是陈麒宇，目前为中国科学院自动化研究所模式识别与智能系统专业直博生，导师为张正涛研究员。

🌐 Welcome to my homepage! I am Qiyu Chen, a direct Ph.D. student in Pattern Recognition and Intelligent Systems at the Institute of Automation, Chinese Academy of Sciences (CASIA), supervised by Prof. Zhengtao Zhang.

🔍 我的研究方向聚焦于工业视觉异常检测，尤其关注异常数据合成、零样本异常检测、语义分割、多模态大模型以及面向工业检测的鲁棒计算机视觉系统。

🔍 My research focuses on industrial visual anomaly detection, especially anomaly data synthesis, zero-shot anomaly detection, semantic segmentation, multimodal large models, and robust computer vision systems for industrial inspection.

😊 如果您对相关研究合作或学术交流感兴趣，欢迎随时与我联系。

😊 Please feel free to contact me if you are interested in research collaboration or academic discussion.

Resume

News

2026.03: 🎉🎉 Our paper CoPS: Conditional Prompt Synthesis for Zero-Shot Anomaly Detection was accepted by CVPR Findings; the code has been open-sourced.
2026.01: 🎉🎉 Our paper MRAD: Zero-Shot Anomaly Detection With Memory-Driven Retrieval was accepted by ICLR; the code has been open-sourced.
2025.11: 🎉🎉 Our paper TDSS: Task Dynamic-Synergistic Skill Adaptation for Efficient and Scalable Multi-Task Learning was accepted by AAAI.
2025.11: 🎉🎉 Our paper Parameter-, Memory-, Time-Efficient Multi-Task Dense Vision Adaptation was accepted by AAAI.
2025.11: 🎉🎉 Our paper Anomagic: Crossmodal Prompt-driven Zero-shot Anomaly Generation was accepted by AAAI; the code has been open-sourced.
2025.05: 🎉🎉 Our paper Center-aware Residual Anomaly Synthesis for Multiclass Industrial Anomaly Detection was published in IEEE Transactions on Industrial Informatics; the code has been open-sourced.
2025.03: 🎉🎉 Our paper Bayesian Prompt Flow Learning for Zero-Shot Anomaly Detection was accepted by CVPR; the code has been open-sourced.
2024.10: 🎉🎉 Our paper Progressive Boundary Guided Anomaly Synthesis for Industrial Anomaly Detection was published in IEEE Transactions on Circuits and Systems for Video Technology; the code has been open-sourced.
2024.07: 🎉🎉 Our paper A Unified Anomaly Synthesis Strategy with Gradient Ascent for Industrial AD and Localization was accepted by ECCV; the code has been open-sourced.

Publications

CoPS: Conditional Prompt Synthesis for Zero-Shot Anomaly Detection [Paper] [Code]
Qiyu Chen, Zhen Qu, Wei Luo, Haiming Yao, Yunkang Cao, et al.
IEEE/CVF Conference on Computer Vision and Pattern Recognition Findings (CVPR-F).
Center-aware Residual Anomaly Synthesis for Multiclass Industrial Anomaly Detection [Paper] [Code]
Qiyu Chen, Huiyuan Luo, Haiming Yao, Wei Luo, Zhen Qu, et al.
IEEE Transactions on Industrial Informatics (TII).
Progressive Boundary Guided Anomaly Synthesis for Industrial Anomaly Detection [Paper] [Code]
Qiyu Chen, Huiyuan Luo, Han Gao, Chengkan Lv, Zhengtao Zhang
IEEE Transactions on Circuits and Systems for Video Technology (TCSVT).
A Unified Anomaly Synthesis Strategy with Gradient Ascent for Industrial AD and Localization [Paper] [Code]
Qiyu Chen, Huiyuan Luo, Chengkan Lv, Zhengtao Zhang
European Conference on Computer Vision (ECCV).
TDSS: Task Dynamic-Synergistic Skill Adaptation for Efficient and Scalable Multi-Task Learning [Paper]
Haiming Yao, Qiyu Chen, Wei Luo, Zheng Zhang, Jianxing Liao, et al.
AAAI Conference on Artificial Intelligence (AAAI).
MRAD: Zero-Shot Anomaly Detection With Memory-Driven Retrieval [Paper] [Code]
Chaoran Xu, Chengkan Lv, Qiyu Chen, Feng Zhang, Zhengtao Zhang
International Conference on Learning Representations (ICLR).
Parameter-, Memory-, Time-Efficient Multi-Task Dense Vision Adaptation [Paper]
Haiming Yao, Wei Luo, Qiyu Chen, Jianxing Liao, Wei You
AAAI Conference on Artificial Intelligence (AAAI).
Anomagic: Crossmodal Prompt-driven Zero-shot Anomaly Generation [Paper] [Code]
Yuxin Jiang, Wei Luo, Hui Zhang, Qiyu Chen, Haiming Yao, et al.
AAAI Conference on Artificial Intelligence (AAAI).
Bayesian Prompt Flow Learning for Zero-Shot Anomaly Detection [Paper] [Code]
Zhen Qu, Xian Tao, Xinyi Gong, Shichen Qu, Qiyu Chen, et al.
IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR).
Distributed Integrated Intelligent Pot Using Remote Monitor and Control System [Paper]
Qiyu Chen, Tianming Li, Shengyue Wang, Mingzhong Huang
Computer Science and Application (CSA).
The Remote Intelligent Virtual Reality Monitoring and Control System for Flower Maintenance [Paper]
Qiyu Chen, Peng Gong, Renchun Guo, Gaojian Zhang
Computer Science and Application (CSA).

Projects

Low-contrast fabric anomaly segmentation

Unsupervised Image Segmentation for Low-Contrast Anomalies

Developed fabric defect detection and data acquisition equipment for a university–industry collaboration project with Weiqiao Textile.
Achieved 99.9% domain-level SOTA performance for subtle fabric defects.
Proposed two industrial defect datasets: WFDD in the ECCV paper and ITDD in the IEEE TII paper.

Patents

Remote Intelligent Flower-Watering System Based on Virtual Reality Technology, CN105532390B.
Multi-Point Sampling Device and UAV Carrying the Multi-Point Sampling Device, CN111551401B.
Intelligent Flowerpot Control System, Flowerpot, and Control Method, CN111657003B.
Defect Detection Method, System, Device, and Computer-Readable Storage Medium, CN118279280B.
Fabric Surface Defect Detection Method and Related Apparatus, CN119810058B.

Awards

First Prize, China Intelligent Robot Fighting Competition.
Second Prize, Chinese Collegiate Computing Design Competition.
Third Prize, National University Intelligent Car Race.
First Prize, CAS Industrial Vision Achievement Scholarship.
Second Prize, CASIA Climbing Scholarship.
Merit Student, University of Chinese Academy of Sciences.
Outstanding Graduate, Tongji University.
Vice President, College of Electronics and Information Engineering, Tongji University.

Education

Sep. 2021 - Jul. 2026, Institute of Automation, Chinese Academy of Sciences

Direct Ph.D. Student in Pattern Recognition and Intelligent Systems, Beijing, China
Advisor: Prof. Zhengtao Zhang
Sep. 2017 - Jul. 2021, Tongji University

B.Eng. in Communication Engineering, College of Electronics and Information Engineering, Shanghai, China

Qiyu Chen