Huimin Zeng

Hi there! 🙌 I'm Huimin Zeng (曾慧敏), a second-year Ph.D. student at SmileLab, Northeastern University, advised by Prof. Raymond Fu. Before joining Northeastern, I received my M.S. degree from the University of Science and Technology of China (USTC) in 2024 under the supervision of Prof. Zhiwei Xiong, and my B.S. degree from Ocean University of China (OUC) in 2021. I also interned at Sanofi (2025), where I was fortunate to collaborate with inspiring researchers in industry.

Email  /  LinkedIn  /  WeChat  /  GitHub  /  Google Scholar

profile photo

News

[Mar. 2026]  One paper for HDR novel view synthesis is accepted to CVPR 2026.

[Dec. 2025]  Passed my PhD qualifying exam, thanks to my advisor and committee members! Cheers! 🎉

[Nov. 2025]  One paper for 3D super-resolution is accepted to AAAI 2026.

[Sep. 2024]  Begin my journey at Northeastern University.

Research

My research focuses on adaptive and interpretable computational photography, with a particular emphasis on 3D scene reconstruction, MLLMs, and low-level vision. I'm also open to discussions and collaborations. 🤝 Feel free to reach out!

project image

Physically Inspired Gaussian Splatting for HDR Novel View Synthesis


Huimin Zeng, Yue Bai, Hailing Wang, Yun Fu
CVPR 2026
paper / code / website

We present PhysHDR-GS, a physically inspired HDR novel view synthesis framework that models scene appearance via intrinsic reflectance and adjustable ambient illumination, thus modeling appearance-related dynamic details.
project image

Arbitrary-Scale 3D Gaussian Super-Resolution


Huimin Zeng, Yue Bai, Yun Fu
AAAI 2026
paper / arxiv / code / website

We make the first attempt to achieve 3D super-resolution of arbitrary scale factors with a single 3DGS model, providing a unified and efficient solution for flexible, high-resolution novel view synthesis.
project image

Plug-and-Play Versatile Compressed Video Enhancement


Huimin Zeng, Jiacheng Li, Zhiwei Xiong
CVPR 2025
paper / arxiv / video / code / website

We introduce a versatile quality enhancement framework that leverages bitstream information to adaptively enhance videos of different compression levels and assists various downstream vision tasks.
project image

All-in-One Image Compression and Restoration


Huimin Zeng, Jiacheng Li, Ziqiang Zheng, Zhiwei Xiong
WACV 2025 (Oral Presentation)
paper / arxiv / code

We design the first unified pipeline for all-in-one image compression and restoration, which equips the neural image codec with the restoration capability and improves its generalization ability against various degradations.
project image

MarineInst: A Foundation Model for Marine Image Analysis with Instance Visual Description


Ziqiang Zheng, Yiwei Chen, Huimin Zeng, Tuan-Anh Vu, Binh-Son Hua, Sai-Kit Yeung
ECCV 2024 (Oral Presentation)
paper / code / website

We propose MarineInst, a powerful and flexible marine foundation model, which could perform the instance visual description task in an automatic or interactive manner.

Design and source code from Jon Barron's website