‹

›

Hi, I am Yudi Xie (谢宇迪), a third-year undergraduate student in Software Engineering (Honor Class) at Wuhan University. I am currently a student at the Multimedia Analysis and Reasoning Lab (MARS Lab), advised by Prof. Mang Ye. I plan to continue my research with Prof. Anyi Rao at MMLab@HKUST.

My research interests focus on Computer Vision, Multi-Modal Learning, Videogen and World model.

Beyond research, I enjoy orienteering, filming vlogs, marathon running, photography, geocaching, fingerstyle guitar, table tennis, singing, and more.

xieyudiwhu AT 163.com / GitHub / Blog / Music / Travel

Show all works

View-Spectral Reconciliation Learning for Text-based Multispectral Aerial-Ground Person Re-Identification

Advances in Neural Information Processing Systems (NeurIPS) 2026

Zhongao Zhou*, Bin Yang*, Yuxuan Zhao, Yudi Xie, Jun Chen†, Mang Ye†

We introduce text-based multispectral aerial-ground person re-identification and the TMAG dataset. Our View-Spectral Reconciliation Learning framework uses textual guidance to align pedestrian representations across viewpoints and spectral modalities.

Towards Cross-Modal Preservation, Consistency and Alignment for Privacy-Preserving Visible-Infrared Person Re-Identification

IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR) 2026

Yudi Xie, Zhongao Zhou, Bin Yang†, Zhenghan Chen, Mang Ye†

This work studies privacy-preserving visible-infrared person re-identification through cross-modal preservation, consistency, and alignment, aiming to preserve privacy while enabling cross-modal retrieval.

Paper Project

Academic Services:

• Student Committee Member, AI Film Frontiers (CVEU), SIGGRAPH Asia 2026
• Reviewer, IEEE Transactions on Pattern Analysis and Machine Intelligence (TPAMI)