I am Changhao He (何长浩), a second-year Master’s student at the College of Computer Science, Sichuan University, supervised by Prof. Peng Hu (胡鹏). Starting this year, I will continue as a Ph.D. student under his guidance. My research focuses on Multi-modal Learning and Noisy Correspondence.
Our group maintains the Awesome Noisy Correspondence repository, a resource hub for research in noisy correspondence. If you have any suggestions or ideas, we warmly invite you to get in touch with us.
🔥 News
- 2025.02.27: 🎉🎉 one paper was accepted by IEEE/CVF Computer Vision and Pattern Recognition (CVPR 2025)! Thanks to all coauthors.
- 2024.07.16: 🎉🎉 one paper was accepted by ACM Multimedia (ACM MM 2024)! Thanks to all coauthors.
📝 Publications
(#: Equal contribution, †: Corresponding author )
[CVPR’25] Learning with Noisy Triplet Correspondence for Composed Image Retrieval
Shuxian Li#, Changhao He#, XitingLiu, Joey Tianyi Zhou, Xi Peng, Peng Hu†
- Introducing a novel setting (learning with noisy triplet correspondence) in CIR, thus offering a new design perspective for existing supervised methods.
- Reframing triplet noise as an adapter mismatch problem and using visual variation modeling to align reference-target modifications.
[ACM MM’24] Robust Variational Contrastive Learning for Partially View-unaligned Clustering
Changhao He, Hongyuan Zhu, Peng Hu†, Xi Peng
- Propose a novel Variational Contrastive Learning paradigm to learn view-common representation and view-specific one simultaneously.
- Design a robust inter-view contrastive loss to optimize clean and noisy pairs selectively.
🎖 Honors and Awards
- 2024.11 National Scholarship
- 2024.10 Outstanding Student of Sichuan University
📖 Educations
- 2023.09 - (now), Pursuing M.S. degree at the College of Computer Science, Sichuan University.
- 2019.09 - 2023.06, B.S. degree at the College of Mechanical Engineering, Sichuan University.