๐๐ป Welcome!
Hello! I am Jiaqi Liu (ๅๅฎถ็ฆ), a third-year undergraduate student at the School of Computer Science, Wuhan University.
Currently, I am fortunate to be advised by Prof. Mang Ye at the MARS Lab. I also have the great opportunity to collaborate on a research project with HKUST, focusing on visual representation learning for complex real-world data.
๐ I am actively seeking Fall 2027 Ph.D. opportunities in Computer Science! Please feel free to reach out if you are interested in my profile.
๐ Research Interests
My long-term research goal is to build intelligent, adaptable, and multimodal systems that can reason about the real world. Currently, my focus encompasses the following areas:
-
Federated & Graph Learning: Exploring how to efficiently model complex structural data and enable continual, collaborative learning across distributed environments without compromising privacy.
-
Large Language Models & Agents: Building upon my prior research in distributed systems and structured data, I aim to explore how multi-agent collaboration and symbolic knowledge representations can enhance the reasoning and alignment capabilities of Foundation Models.
-
Vision-Language Models (VLMs) & Multimodal Reasoning: Moving beyond linguistic or graph representations, I am eager to explore Vision-Language Models (VLMs) as a new frontier. My goal is to investigate how to build robust multimodal systems that can align visual perception with complex, real-world reasoning.
๐ฅ News
- 2026.02: ย ๐๐ One paper was accepted by CVPR 2026. See you in Denver!
๐ Publications
โ Equal Contribution
FedSDR: Federated Graph Learning with Structural Noise Detection and Reconstruction
Jiaqi Liuโ , Zihan Tanโ , Guancheng Wan, Wenke Huang, He Li, Mang Ye
Highlight Presentation
We propose FedSDR, a spectra-based federated graph learning framework, featuring two key designs: (1) structural noise-aware aggregation for global noise detection and mitigation, and (2) robust local structure reconstruction guided by healthy global knowledge to repair corrupted graphs.
๐ Honors and Awards
2026.05 Lead Investigator, National Key Supported Project, Undergraduate Training Programs for Innovation (Top 4 of 1,245 Univ-wide)
2026.03 Scientific Innovation Pioneer (Sole Recipient among all undergraduate and graduate students in the School)
2025.11 National First Prize in the 19th โChallenge Cupโ Academic and Scientific Works Competition (Top 0.07% Nationwide)
2025.11 Fiberhome Communication Scholarship
2025.09 Outstanding Student Scholarship
2025.05 Lei Jun Computer Innovation and Development Fund
๐ Educations
Undergraduate, Software Engineering, Wuhan University
