Xu Zheng

About Me

I am a Ph.D. candidate in the AI Thrust at The Hong Kong University of Science and Technology, Guangzhou campus. I am fortunate to be advised by Xuming Hu @ HKUST and Raymond Chi-Wing Wong @ HKUST.

I also serve as a Resident Doctoral Researcher at INSAIT, under the supervision of Prof. Luc Van Gool and Dr. Danda Paudel. Recently, I have also been collaborating with Linfeng Zhang @ SJTU and Kailun Yang @ HNU.

My research centers on machine perception, reasoning, and interaction with the physical world.

Currently, I focus on:

Multi-modal Learning: foundation models, representation learning, self-supervision, contrastive alignment.
Multi-modal Large Models: spatial & embodied reasoning, debiasing, retrieval-augmented understanding & generation.
Multi-modal in Computer Vision: novel sensors, sensor fusion, scene understanding.

🔥 I am actively seeking job opportunities (academia & industry) for Fall 2026!

Latest News

2025.06: One paper accepted to BMVC 2025!

2025.06: Four papers accepted to ICCV 2025!

2025.06: Our paper is selected as Best Paper at CVPR 2025 @ TMM Open-World! Code and Paper

2025.06: One paper accepted to IROS 2025! Code and Paper

2025.05: Two paper accepted to ACL 2025 Findings! Paper1 and Paper2

2025.05: One paper accepted to ICML 2025! Paper

2025.04: The first RAG in CV survey released! Code and Paper

2025.04: Our paper accepted to CVPR 2025 @ TMM Open-World as Oral Presentation! Code and Paper

2025.02: Visit INSAIT as a Resident Doctoral Researcher! LinkedIn

2025.01: Successfully passed PhD Qualifying Examination!

2024.12: Invited as an Area Chair of PDLM @ AAAI 2025.

2024.10: One paper accepted to IEEE TPAMI

2024.10: Oral presentation @ ECCV 2024 Oral Session 5A: Segmentation [video]

2024.09: One paper accepted to Pattern Recognition

2024.07: Three papers (one Oral (1.5%)) accepted to ECCV 2024

2024.03: One paper accepted to IEEE CAI 2024

2024.03: One paper accepted to Pattern Recognition

2024.03: Five papers (one Highlight (2.8%)) accepted to CVPR 2024

2024.02: Two papers accepted to ICRA 2024

2023.07: Two papers accepted to ICCV 2023

2023.03: One paper accepted to CVPR 2023

Talks

"Retrieval-augmented Realistic Image Generation via Self-reflective Contrastive Learning"
VIVO (Shenzhen) August, 2025. Invited talk by Dr. Kanzhi Wu at VIVO, China, August 2025.

Mentorship

I am honored to mentor the following students and researchers:

Current:

Yuanhuiyi Lyu, Ph.D. @ HKUST(GZ)
Lutao Jiang, Ph.D. @ HKUST(GZ)
Jialei Chen, Ph.D. @ Nagoya University
Mengzhen Chi, Ph.D. @ NEU
Zihao Dongfang, Research Assistant @ HKUST(GZ)
Chenfei Liao, M.Phil. @ HKUST(GZ)
Junha Moon (문준하), M.Phil. @ HKUST(GZ)
Ziqiao Weng, Undergraduate @ SCU
Yulong Guo, Master’s @ ZJU
Kaiyu Lei, Undergraduate @ XJTU
Zhenquan Zhang, M.Phil. @ SCUT
Boyuan Zheng, M.Phil. @ Tongji University
Zijie Lin, Master’s @ USTC
Leyi Sheng, Undergraduate @ HKUST(GZ)

Past:

Ding Zhong, Master’s @ University of Michigan
Zhengxuan Jiang, M.Phil. @ ZJU
Yunhao Luo, Student Researcher @ Georgia Tech
Tianbo Pan, Incoming Ph.D. @ NUS

✉️ Feel free to contact me for discussion and collaboration!

Selected Publications

ICCV 2025

Xu Zheng, et al., Reducing unimodal bias in multi-modal semantic segmentation with multi-scale functional entropy regularization.

ICCV 2025

Ding Zhong*, Xu Zheng*, et al., Omnisam: Omnidirectional segment anything model for uda in panoramic semantic segmentation.

TPAMI 2024

Xu Zheng, et al., 360sfuda++: Towards Source-free UDA for Panoramic Segmentation by Learning Reliable Category Prototypes.

PR 2025

Xu Zheng, et al., Distilling Efficient Vision Transformers from CNNs for Semantic Segmentation.

ECCV 2024 (Oral, 1.5%)

Xu Zheng, et al., Learning Modality-Agnostic Representation for Semantic Segmentation.

ECCV 2024

Xu Zheng, et al., Centering the Value of Every Modality: Towards Efficient and Resilient Modality-Agnostic Semantic Segmentation.

CVPR 2024

Xu Zheng, et al., EventDance: Unsupervised Source-Free Cross-Modal Adaptation for Event-Based Object Recognition.

CVPR 2024

Yuanhuiyi Lyu*, Xu Zheng*, et al., UniBind: LLM-Augmented Unified and Balanced Representation Space to Bind Them All.

CVPR 2024

Xu Zheng, et al., Semantics, Distortion, and Style Matter: Towards Source-Free UDA for Panoramic Segmentation.

ICRA 2024

Xu Zheng, et al., Transformer-CNN Cohort: Semi-Supervised Semantic Segmentation by the Best of Both Students.

ICCV 2023

Xu Zheng, et al., Look at the Neighbor: Distortion-Aware Unsupervised Domain Adaptation for Panoramic Semantic Segmentation.

CVPR 2023

Xu Zheng, et al., Both Style and Distortion Matter: Dual-Path Unsupervised Domain Adaptation for Panoramic Semantic Segmentation.

Research Experience

[Feb. 2025 – Present] Resident Doctoral Researcher, INSAIT, Sofia University
Supervisor: Prof. Luc Van Gool, Dr. Danda Paudel
[Oct. 2021 – Oct. 2024] Research Assistant, VLISLAB, HKUST
Supervisor: Addison Lin Wang
[Aug. 2022 – Sep. 2022] Visiting Researcher, Nagoya University (Summer Program)

Services

Area Chair: PDLM Workshop @ AAAI 2025
Reviewers: IJCV, TIP, TNNLS, TMM, TCI, Neurocomputing, etc.
PC Members: ICLR (2024,2025), CVPR (2025), ICML (2025), ECCV (2024), NeurIPS (2024), ACM MM (2025), ICRA (2025), ICME (2025)

Honors & Awards

2025.06 🏆 Best Paper, CVPR 2025 Workshop – “Navigating the Future: Ensuring Trustworthiness in Multi-Modal Open-World Intelligence”
2024.08 💰 HKUST Postgraduate Scholarship
2023.08 💰 HKUST Postgraduate Scholarship
2023.07 🏆 Outstanding Master’s Thesis, Liaoning Province / Northeastern University (Top 1%)
2022.09 💰 National Scholarship of China (Top 2%)
2020–2023.09 💰 First-Class Master Scholarship, Northeastern University (Top 20%)
2019.06 🏆 Outstanding Graduate, Liaoning Province / Northeastern University (Top 2%)