Biography

Haoyu Zhao is a final-year Ph.D. student at the Institute of Trustworthy Embodied AI (复旦大学可信具身智能研究院), Fudan University (QS 30), and a member of Fudan Vision and Learning Lab (复旦大学视觉与学习实验室,FVL), under the supervision of Prof. Zuxuan Wu and Prof. Yu-Gang Jiang.

Since 2019, his research has focused on AIGC and Embodied AI, with particular interests in controllable diffusion models, world models, and spatial intelligence.

He has published over twenty papers in top-tier venues, including ICML, ECCV, CVPR, AAAI, ICLR, TMM, and FCS.

📝 Selected Publications

  • Haoyu Zhao, Jiaxi Gu, et al. CameraNoise: Enabling Faithful Camera Control in Video Diffusion through Geometry-Flow-Guided Noise Warping. ICML, 2026.

  • Haoyu Zhao, Yuang Zhang, et al. DCDM: Divide-and-conquer diffusion models for consistency-preserving video generation. 🏆 AAAI’26 CVM Challenging Top Team, 2026. (Media Coverage: 智猩猩AI)

  • Qingping Zheng, Bo Huang, Yang Liu, Haoyu Zhao, et al. ReFocusEraser: Refocusing for small object removal with robust context-shadow repair. ICLR, 2026.

  • Haoyu Zhao, Jiaxi Gu, et al. LSTD: Long short-term temporal diffusion for video generation. TMM, 2025. (Media Coverage: 机器之心)

  • Zihao Zhang, Haoran Chen, Haoyu Zhao, et al. Eden: Enhanced diffusion for high-quality large-motion video frame interpolation. CVPR, 2025.

  • Cong Wang, Panwen Hu, Haoyu Zhao, et al. Uniadapter: All-in-one control for flexible video generation. TCSVT, 2025.

  • Haoyu Zhao, Tianyi Lu, et al. MagDiff: Multi-alignment diffusion for high-fidelity video generation and editing. ECCV, 2024.

  • Haoyu Zhao, Weidong Min, et al. Scene-adaptive crowd counting method based on meta learning with dual-input network DMNet. Frontiers of Computer Science (FCS), 2023.

  • Haoyu Zhao, Weidong Min, et al. Memory-efficient document layout analysis method using LD-net. Multimedia Tools and Applications (MTAP), 2023.

  • Qi Wang, Yuling Zhong, Weidong Min, Haoyu Zhao, et al. Dual similarity pre-training and domain difference encouragement learning for vehicle re-identification in the wild. PR, 2023.

  • Haoyu Zhao, Qi Wang, et al. Need Only One More Point (NOOMP): Perspective adaptation crowd counting in complex scenes. TMM, 2022.

  • Zitai Wei, Weidong Min, Qi Wang, Qian Liu, Haoyu Zhao. ECNFP: Edge-constrained network using a feature pyramid for image inpainting. Expert Systems with Applications (ESWA), 2022.

  • Haoyu Zhao, Weidong Min, et al. SPACE: Finding key-speaker in complex multi-person scenes. IEEE Transactions on Emerging Topics in Computing (TETC), 2021.

  • Haoyu Zhao, Weidong Min, et al. MSR-FAN: Multi-scale residual feature-aware network for crowd counting. IET IP, 2021.

  • Haoyu Zhao, Weidong Min, et al. Illumination-enhanced crowd counting based on IC-Net in low lighting conditions. ICIG, 2021.

  • Qi Wang, Weidong Min, Qing Han, Qian Liu, Cheng Zha, Haoyu Zhao, et al. Inter-domain adaptation label for data augmentation in vehicle re-identification. TMM, 2021.

🎖 Honors and Awards

  • AAAI’26 CVM Top Team (Role: PI), 2026

  • Outstanding Student Award, Fudan University, 2024

  • Academic Scholarship (First Class), Fudan University, awarded multiple times, 2022-2026

  • National Scholarship, Ministry of Education, China (Top 1%), 2022

  • Outstanding Graduate, Nanchang University, China, 2022

  • Academic Scholarship (First Class), Nanchang University, awarded multiple times, 2019-22

📖 Educations

  • 2022.09 - Present, Ph.D in College of Computer Science and Artificial Intelligence, Fudan University. Supervisor: Prof. Yu-Gang Jiang and Prof. Zuxuan Wu.
  • 2019.09 - 2022.06, M.S. in School of Mathematics and Computer Sciences, Nanchang University. Supervisor: Prof. Weidong Min.
  • 2015.09 - 2019.06, B.S. in School of Information Engineering, Nanchang University.

💼 Work Experience

  • 2025.05 - Present, Research Intern (Qingyun Talent Program), Advertising Marketing Services (AMS), Tencent. Conducted research on digital human and contributed to their deployment in Tencent Ads Miaosi.

  • 2023.03 - 2025.03, Research Intern, Huawei Noah’s Ark Lab, Shanghai. Worked on AIGC research and its deployment in Huawei Mate-series phones.

📋 Academic Services

Reviewer for IJCV, TIP, TMM, KBS, TCSVT, NeurIPS, ICLR, ICML, CVPR, AAAI, ACM MM, etc.


Updated on May 1, 2026.

Visitor Map

Live tracking enabled in Apr. 2026.