I am a Ph.D. student at CVLAB, University of Electronic Science and Technology of China (UESTC), supervised by Professor Mao Ye. Specifically, I am enrolled in a combined Master and Doctoral program, with my Master's study spanning from 2020 to 2022 and my Ph.D. research continuing from 2022 to the present. Before this, I completed my Bachelor's degree at UESTC (2016–2020). My primary research interests include Large Cloud Model Adaptation and Domain Adaptive Object Detection. Additionally, I have a broad interest in language-vision models, encompassing recognition, detection, and segmentation tasks.
I am open to research collaborations. If you share similar interests or are interested in my previous work, please feel free to contact me.
💥 News
- [2025/03] 1 paper about ODE-flow based image generation is accepted by ESWA! Congrats to Jian Yue!
- [2025/02] 2 papers are accepted by CVPR 2025! Congrats to Lihua and Liting!
- [2024/09] 1 paper that explores cloud object detector adaptation is accepted by NeurIPS 2024!
- [2024/08] 1 paper about black-box domain adaptation is accepted by ACMM MM 2024! Congrats to Siying!
- [2023/04] 1 paper using mutual learning for UDA is accepted by TCSVT! Congrats to Lihua!
- [2022/08] 1 paper about unsupervised domain adaptation (UDA) is accepted by ACMM MM 2022! Congrats to Lihua!
- [2022/03] 1 paper about source-free object detection by overlooking domain style is accepted by CVPR 2022 (ORAL)!
- [2021/12] 1 paper about source-free object detection is accepted by ACMM MM Asia 2021! Congrats to Dan Zhang!
🎯 Publications
Shuaifeng Li, Mao Ye, Lihua Zhou, Nianxin Li, Siying Xiao, Song Tang, Xiatian Zhu. The Thirty-eighth Annual Conference on Neural Information Processing Systems (NeurIPS), 2024.
Key Words: Cloud Object Detector Adaptation; Gradient Direction Alignment
We propose to explore an interesting and promising problem CODA.
Paper | Code | Project | Slides | Poster | Blog | Zhihu
Shuaifeng Li, Mao Ye, Xiatian Zhu, Lihua Zhou, Lin Xiong. Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), 2022, ORAL.
Key Words: Source-Free Object Detection; Style Enhancement; Overlooking Style
We propose a novel learning to overlook domain style strategy for SFOD.
Paper | Code | Project | Video | Slides | Poster | Summary
Siying Xiao, Mao Ye, Qichen He, Shuaifeng Li, Song Tang, Xiatian Zhu. Proceedings of the 32nd ACM International Conference on Multimedia (ACM MM), 2024.
Key Words: Black-box Domain Adaptation; CLIP; Adversarial Experts
Paper | Code | Video
Jiuzheng Yang, Song Tang, Yangkuiyi Zhang, Shuaifeng Li, Mao Ye, Jianwei Zhang, Xiatian Zhu. arXiv preprint, 2024.
Key Words: Source-Free Object Detection; Weak-to-Strong Contrastive Learning
Paper
Lihua Zhou, Siying Xiao, Mao Ye, Xiatian Zhu, Shuaifeng Li. IEEE Transactions on Circuits and Systems for Video Technology (TCSVT), 2023.
Key Words: Unsupervised Domain Adaptation; Adaptive Mutual Learning
Paper
Lihua Zhou, Mao Ye, Xiatian Zhu, Shuaifeng Li, Yiguang Liu. Proceedings of the 30th ACM International Conference on Multimedia (ACM MM), 2022.
Key Words: Unsupervised Domain Adaptation; Class Discriminative Adversarial Learning
Paper | Code | Video
Dan Zhang, Mao Ye, Lin Xiong, Shuaifeng Li, Xue Li. Proceedings of the 3rd ACM International Conference on Multimedia in Asia (ACM MM Asia), 2021.
Key Words: Source-Free Object Detection; Batch Normalization Adaptation
Paper
🏆️ Selected awards
- [2024] Second Prize, China Postgraduate Mathematical Contest in Modeling
- [2020–2024] Multiple Academic Excellence Scholarships (First Class, Second Class)
- [2022] Academic Seedling Award, University of Electronic Science and Technology of China
- [2016–2020] Multiple National Encouragement Scholarships and Outstanding Student Scholarships (Top 20% Students)
🧭 Service
- Conference Reviewer
- ACM International Conference on Multimedia (ACM MM), 2023, 2024, 2025
- The IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), 2025
- International Conference on Computer Vision, ICCV 2025
- Journal Reviewer
- Knowledge-Based Systems (KBS)