Chen Xu

Ph.D. Student, AIGC-DL Lab, University of Macau

Research Interests

• Computer Vision, with an emphasis on Parameter-Efficient Fine-Tuning (PEFT) of large models.
• Diffusion and Flow-Matching Acceleration for image generation, image restoration, and vision-language-action (VLA) models.

Academic Qualifications

Ph.D. Student | University of Macau | Macau

AIGC-DL Lab | Advisor: Prof. Zhedong Zheng

Bachelor of Engineering in Computer Science and Technology | Nankai University

NKICS Lab | Advisors: Prof. Tao Li and Prof. Kai Wang

Past Experience

Research Assistant | AIGC-DL Lab, University of Macau | Mar 2026 - Aug 2026

Institute of Machine Intelligence | Nankai University | Sep 2023 - Jun 2025

• Conducted research on TS-SAM, Resfusion, and MOST under Prof. Kai Wang and Prof. Tao Li; contributed to model design, training, evaluation, dataset development, and paper writing.

Undergraduate Innovation and Entrepreneurship Project | Nankai University | Apr 2023 - Apr 2024

• Led a competitive-sports video tracking project; generated pseudo-labels with SAM, curated a custom dataset, proposed MOST, and designed the lightweight MOST-SAM module.

AI Class Trainee | Chinasoft International Internship and Training | Jun 2023 - Jul 2023

• Led the Yellowbook game-forum project; built a local-server platform and trained a textRNN classifier with about 87% accuracy.

Research Paper Project | Yau Science Award Mathematics | Apr 2020 - Nov 2020

• Extended random interval graphs to higher-dimensional spaces and wrote “Research on Random Interval Graphs and Their Generalized Properties.”

Awards

• 2024	K Award (Participant), American Collegiate Mathematical Modeling Contest

• 2024	Team Instructor, U.S. Collegiate Mathematical Modeling Contest

• 2024	Third Prize, Operating System Competition, Computer System Development Capability Competition

• 2023	Fourth Place, Doubles, 2023 Computer Science College Dean’s Cup Badminton Tournament

• 2023	Honorable Team, Chinasoft International AI Class Internship and Training

• 2022	Bronze Award, Tianjin Volunteer Service Competition

• 2022	Outstanding Team, National Social Practice to the Countryside

• 2022	Outstanding Social Practice Field Trip Report

• 2020	First Prize, Northern Division; Finalist Award, Yau Science Award Mathematics

• 2020	Outstanding Student, Shing-Tung Yau International Summer Math Camp for Secondary School Students

Publications

1. ScaleResfusion: Residual Rectified Flow based on Residual Vector Field.
   arXiv preprint arXiv:2607.25275, 2026. Paper | Code
   Zhenning Shi^*, Chen Xu*, Junhao Zhang, Kefei Zhang, Linjie Liu, Zhedong Zheng, Tao Li.

2. Resfusion: Denoising Diffusion Probabilistic Models for Image Restoration Based on Prior Residual Noise.
   Advances in Neural Information Processing Systems 37 (NeurIPS), 2024, pp. 130664-130693. Paper
   Zhenning Shi, Haoshuai Zheng, Chen Xu, Changsheng Dong, Bin Pan, Xueshuo Xie, Along He, Tao Li, Huazhu Fu.

3. TS-SAM: Fine-Tuning Segment-Anything Model for Downstream Tasks.
   2024 IEEE International Conference on Multimedia and Expo (ICME), 2024, pp. 1-6. Paper
   Yang Yu, Chen Xu, Kai Wang.

^* Co-first authors (equal contribution).