💫 About Me
Hi there!👋
Welcome to the homepage of Yushu Li (李昱澍)! I am a third-year master’s student at the Geometric Perception and Intelligence Research Lab (Gorilla Lab), South China University of Technology (SCUT), where I am supervised by Prof. Kui Jia. I also closely collaborate with Dr. Xun Xu. Previously, I was a visiting student at the Institute for Infocomm Research (I2R) under the Agency for Science, Technology and Research (A*STAR), Singapore, where I was also supervised by Dr. Xun Xu. I completed my B.Eng. at the School of Electronic and Information Engineering, SCUT, in June 2022.
My research focuses on test-time adaptation, efficient tuning, and vision-language models. I am always open to collaborations with researchers and engineers. If you’re interested in my work or have any inquiries, feel free to reach out via email at yushuliysl@gmail.com or eeyushuli@mail.scut.edu.cn.
🌟 I am actively looking for a Ph.D. position starting from Fall 2025. If you are interested in my profile, please feel free to reach out to me.
🔥 News
- 2024.09: ⛵️ I am joining the Shanghai Artificial Intelligence Laboratory for research intern in Shanghai, China.
- 2024.09: 🎓 I have completed my SIPGA program at I2R, A*STAR, Singapore.
- 2024.08:🎉 I will be serving as a reviewer for ICLR 2025.
- 2024.05: ✨ New preprint on Exploring Human-in-the-Loop Test-Time Adaptation by Synergizing Active Learning and Model Selection is available on arXiv.
- 2024.03: ⛵️ Starting my visiting student program “Singapore Inernational Pre-graduate Award (SIPGA)” at I2R, A*STAR, Singapore.
- 2023.07: 🎉 Our work on On the Robustness of Open-World Test-Time Training: Self-Training with Dynamic Prototype Expansion was accepted by ICCV 2023 as Oral presentation (1.8% acceptance rate).
📝 Publications
Exploring Human-in-the-Loop Test-Time Adaptation by Synergizing Active Learning and Model Selection
Yushu Li*, Yongyi Su*, Xulei Yang, Kui Jia, Xun Xu
arXiv preprint arXiv:2405.18911, 2024.
-
The first work to explore human-in-the-loop by synergizing active learning and model selection for test-time adaptation.
-
A novel TTA model selection method preventing from overfitting to a small validation set and local distribution.
-
A novel test-time selection strategy for active learning to select the most informative samples for human annotation.
On the Robustness of Open-World Test-Time Training: Self-Training with Dynamic Prototype Expansion
Yushu Li, Xun Xu, Yongyi Su and Kui Jia
IEEE/CVF International Conference on Computer Vision (ICCV), 2023 (Oral presentation 1.8% acceptance rate)
-
A new TTT protocol: open-world test-time training (OWTTT).
-
A new TTT method for OWTTT: self-training with dynamic prototype expansion.
-
A new TTT benchmark: evaluating TTT. methods under open-world settings
Dynamic Shuffle: An Efficient Channel Mixture Method
Kaijun Gong, Zhuowen Yin, Yushu Li, Kailing Guo, Xiangmin Xu
arXiv preprint arXiv:2310.02776.
- A novel method for efficient channel mixture to improve the performance of group convolution.
💻 Internships
- 2024.09 - Present, Research Intern, Shanghai Artificial Intelligence Laboratory, Shanghai, China. Mentor: Dr. Peng Gao.
- 2024.03 - 2024.09, Visiting Student, Institute for Infocomm Research (I2R), Agency for Science, Technology and Research (A*STAR), Singapore. Mentor: Dr. Xun Xu.
- 2023.07 - 2024.02, DexForce Technology, Research Intern. Mentor: Prof. Kui Jia.
🎖 Honors and Awards
- First-class Scholarship of South China University of Technology (2020) (Top 4%)
- Second-class Scholarship of South China University of Technology (2019, 2021) (Top 10%)
🎓 Educations
- 2022.09 - 2025.06, M.Eng. in Information and Communication Engineering, South China University of Technology (SCUT), China
- 2018.09 - 2022.06, B.Eng. in Information Engineering, Chu Kochen Honors College, South China University of Technology (SCUT), China
📖 Academic Service
- Conference Reviewer: ACM MM 2024, NeurIPS 2024, ICLR 2025
- Journal Reviewer: IEEE Transactions on Image Processing (TIP)
This page was last modified by Yushu Li on: