Jiaqi Tang

I am a Ph.D. student in Dept. of ECE at Hong Kong University of Science and Technology (HKUST), starting from Fall 2025, supervised by Prof. Qifeng Chen. Prior to this, I earned my M.Phil. in AI at HKUST Guangzhou, jointly supervised by Prof. Ying-Cong Chen and Prof. Qifeng Chen, in 2025. Before that, I obtained B.Eng. in Data Science & Big Data Tech. and Business Administration (Minor) with outstanding graduate at Northwestern Polytechnical University, supervised by Prof. Wei Wei, in 2022. I am working closely with Dr. Xiaogang Xu at MiroMind.

My research focuses on Multimodal Large Language Models (MLLMs), including multimodal reasoning and understanding.

Email / Scholar / Github / LinkedIn / HuggingFace / DBLP / ORCID / ResearchGate / Kaggle / YouTube

Captured at Dolomites, Italy

News

• Nov 2025: One paper is accepted by AAAI2026 (Oral).
• Jun 2025: Two papers are accepted by ICCV2025, including one Highlight.
• May 2025: I am happy to pass my M.Phil. Thesis Defence.
• Apr 2025: One paper is accepted by CVPR2025 Highlight (Top 2.9%).
• Nov 2024: Congratulations! Our paper, "AdaShadow: Responsive Test-time Model Adaptation in Non-stationary Mobile Environments" is selected as Best Paper Honorable Mention in ACM SenSys 2024.
• Sep 2024: One paper is accepted by NeurIPS2024.
• Sep 2024: One paper is accepted by SenSys2024.
• Jul 2024: One paper is accepted by ECCV2024.
• Apr 2024: One survey is accepted by CVPRW2024 (Oral).
• Feb 2024: One paper is accepted by CVPR2024.
• Jul 2023: One paper is accepted by ECAI2023 (Long Oral).

Selective Publications [Full Publication List]

Some representative papers are highlighted.

	Robust-R1: Degradation-Aware Reasoning for Robust Visual Understanding Jiaqi Tang, Jianmin Chen, Wei Wei†, Xiaogang Xu, Runtao Liu, Xiangyu Wu, Qipeng Xie, Jiafei Wu, Lei Zhang, Qifeng Chen† AAAI Conference on Artificial Intelligence (AAAI), 2026 (Oral), *: Equal Contribution bibtex Degradation-aware reasoning for robust visual understanding.
	RhythmGuassian: Repurposing Generalizable Gaussian Model For Remote Physiological Measurement Hao Lu, Yuting Zhang, Jiaqi Tang, Bowen Fu, Wenhang Ge, Wei Wei, Kaishun Wu, Ying-Cong Chen IEEE/CVF International Conference on Computer Vision (ICCV), 2025 (Highlight), *: Equal Contribution bibtex Repurposing generalizable Gaussian model for remote physiological measurement.
	SURGEON: Memory-Adaptive Fully Test-Time Adaptation via Dynamic Sparsity Activation Ke Ma, Jiaqi Tang, Fan Dang, Bin Guo†, Sicong Liu, Cheng Fang, Zhui Zhu, Lei Wu, Ying-Cong Chen, Zhiwen Yu, Yunhao Liu† IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), 2025 (Highlight, Top 2.9%) code / bibtex Memory-adaptive test-time adaptation method that dynamically activates sparsity for efficient adaptation.
	AdaShadow: Responsive Test-time Adaptation for Non-stationary Mobile Environments Cheng Fang, Sicong Liu, Zimu Zhou, Bin Guo†, Jiaqi Tang, Ke Ma, Zhiwen Yu ACM Conference on Embedded Networked Sensor Systems (SenSys), 2024 (Best Paper Honorable Mention, Top 7/313) bibtex Responsive test-time adaptation framework for non-stationary mobile environments.
	Hawk: Learning to Understand Open-World Video Anomalies Jiaqi Tang, Hao Lu, Ruizheng Wu, Xiaogang Xu, Ke Ma, Cheng Fang, Bin Guo, Jiangbo Lu, Qifeng Chen, Ying-Cong Chen† Annual Conference on Neural Information Processing Systems (NeurIPS), 2024 (VALSE poster) code / demo / model / dataset / website / bibtex Learning to understand open-world video anomalies through multi-modal large language models.
	Learning to Remove Wrinkled Transparent Film with Polarized Prior Jiaqi Tang, Ruizheng Wu, Xiaogang Xu, Sixing Hu, Ying-Cong Chen† IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), 2024 code / project page / bibtex Learning-based method to remove wrinkled transparent film using polarized prior information.
	An Incremental Unified Framework for Small Defect Inspection Jiaqi Tang, Hao Lu, Xiaogang Xu, Ruizheng Wu, Sixing Hu, Tong Zhang, Tsz Wa Cheng, Ming Ge, Ying-Cong Chen†, Fugee Tsung 18th European Conference on Computer Vision (ECCV), 2024 code / project page / bibtex An incremental unified framework for small defect inspection in industrial applications.
	High Dynamic Range Image Reconstruction via Deep Explicit Polynomial Curve Estimation Jiaqi Tang, Xiaogang Xu, Sixing Hu, Ying-Cong Chen† 26th European Conference on Artificial Intelligence (ECAI), 2023 (Long Oral) code / arXiv / talk / bibtex High dynamic range image reconstruction through deep explicit polynomial curve estimation.

Internship & Professional Experience

2025 - Now: Research Intern, DeepRoute.ai (Shenzhen, China).
    Exploring VLA Driving Safety.

Summer 2025: Research Intern, Creative AI Lab, Sony (Tokyo, Japan).
    Exploring efficient MLLMs via visual token compression.
    Mentorship: Dr. Hiromi Wakaki.

2024 – 2025: Research Intern, Ovis Group, AI Business Team, Alibaba (Hangzhou, China).
    Exploring preference optimization in the accurate interaction of the GUI agent.
    Exploring multi-modal instruction generation in the Multi-modal Large Language Model.
    Mentorship: Mr. Qing-Guo Chen.

2022 – 2024: Research Intern, SmartMore (Hong Kong SAR).
    Exploring robustness image enhancement algorithms in the industrial environment.
    Mentorship: Dr. Jiangbo Lu and Dr. Sixing Hu.

Awards

Best Paper Honorable Mention (7/313), by ACM SenSys 2024.
Best Intern of the Year, by SmartMore Corporation in 2023.
Outstanding Graduate, by Northwestern Polytechnical University in 2022.
National Scholarship (Top 1/44), by The Ministry of Education of the People's Republic of China in 2021.
Tencent Scholarship - First Class, by Tencent in 2021.
First Class Scholarship, by Northwestern Polytechnical University in 2021.
Winner Award (1st Rank), NTIRE-CVPR (New Trends in Image Restoration and Enhancement) Challenge on Multi-modal Aerial View Object Classification, Track 1 (SAR) in 2021.
Second Class Scholarship, by Northwestern Polytechnical University in 2020.

Professional Activities, Skills & Others

Journal Reviewer:
    • IEEE Transactions on Pattern Analysis and Machine Intelligence (T-PAMI)
    • International Journal of Computer Vision (IJCV)
Conference Reviewer:
    • CVPR (24 - )
    • ICCV (25 - )
    • NeurIPS (24, 25 - )
    • ICML (25 - )
    • ICLR (24, 25 - )
    • ACL Rolling Review (25 - )
    • AAAI (25, 26 - )
    • ECAI (23)
Organization: IEEE Student Member, EurAI Student Member.
Coding: Python (PyTorch, DeepSpeed), Java, C/C++, Matlab, SQL, Verilog, MIPS 32/64, IBM ILOG CPLEX, R and LATEX.
Languages: English (Fluent) and Chinese (Native).
Hobbies: Amateur Go 4 Dan (Certified by Chinese Weiqi Association), Travel, Table Tennis.

Teaching

Fall 2024: AIAA 5023: Foundations of Deep Neural Networks, at HKUST Guangzhou.
Summer 2021: U14M12086S: Introduction of Computer Vision and Image Processing, at Northwestern Polytechnical University.