IMDL-BenCo: A Comprehensive Benchmark for Image Manipulation Detection & Localization
Advances in Neural Information Processing Systems (NeurIPS), Datasets and Benchmarks Track, 2024
周吉喆，四川大学计算机学院副教授， 四川省千人， IEEE TETCI 副主编 J. Imaging 客座主编，主要研究方向为伪造图像检测（Fake Image Detection）及生成与取证统一（Forensic-Embedded Generation）。 我主导开发了 IML-ViT Sparse-ViT IMDL-BenCo ForensicHub 等多个开源项目，更多详情请见 GitHub 与 Google Scholar 主页。
I am an Associate Professor at the College of Computer Science, Sichuan University, and Sichuan Provincial Thousand Talents. I serve as Associate Editor of IEEE TETCI and Guest Editor of J. Imaging. My research focuses on Fake Image Detection and Forensic-Embedded Generation. I have led the development of several open-source projects, including IML-ViT Sparse-ViT IMDL-BenCo ForensicHub. For more details, please visit my GitHub and Google Scholar profiles.
♥ Special Thanks to Ms. Helen Wang. YY
📣 News
May 2026
🏛️ Service Gold Reviewer for ICML 2026.
May 2026
📄 ICML 2026 SICA: First monolithic FID model via semantic-induced constrained adaptation, outperforming 15 state-of-the-art methods.
Apr 2026
🏛️ Service Invited as Guest Editor of the special issue "Recent Advances in Multimedia Imaging and Content Security" for Journal of Imaging.
Apr 2026
🏆 DCIC 2026 Champion (1st / 173 teams) — 数字中国创新大赛 · 数据应用赛道：基于多模态大模型的场景文本图像伪造分析.
Jan 2026
📄 IEEE TMM PASK: Sparse framework for crafting natural adversarial examples against ASR via pitch-aware perturbations.
Jan 2026
📄 AAAI 2026 FGM-HD: Boosting generation diversity of fractal generative models through Hausdorff Dimension induction.
Jan 2026
📄 AAAI 2026 IO-RAE: Pioneering reversible adversarial example framework for audio privacy protection via LLM-guided content obfuscation.
📚 Selected Publications
Pre-training-free Image Manipulation Localization via Non-Mutually Contrastive Learning
IEEE/CVF International Conference on Computer Vision (ICCV), pp. 22346–22356, 2023
ForensicHub: Towards Universal Fake Image Detection and Localization
Advances in Neural Information Processing Systems (NeurIPS), Datasets and Benchmarks Track, 2025
IML-ViT: Benchmarking Image Manipulation Localization by Vision Transformer
arXiv preprint arXiv:2307.14863, 2023
Can We Build a Monolithic Model for Fake Image Detection? SICA: Semantic-Induced Constrained Adaptation for Unified-Yet-Discriminative Artifact Feature Space Reconstruction
arXiv preprint arXiv:2602.06676, 2026
💼 Experience
2022 – Present
Associate Professor
Sichuan University, College of Computer Science
天府峨眉计划
Education
🎓 Students & Alumni
Current Students 在读学生
何雨声
PhD Student
生成与取证统一
杨哲
Master Student
伪造图像检测
王韵菲
Master · Intern @ 腾讯混元
伪造图像检测
Alumni 已毕业学生
张健
Master → 宁德时代
缺陷检测
倪睿洋
Master → 中国电信
因果推理