Junkai Fan

I am a Phd student at the PCALab of Nanjing University of Science and Technology, where I'm fortunate to be advised by Prof. Jian Yang and co-advised by Prof. Jun Li.

Prior to joining NJUST, I obtained my M.S degree in 2019 from Wenzhou University, advised by Prof. Zhengzhou Tang.

Email  /  GitHub  /  Google Scholar  /  WeChat

profile photo
Research

My research is focused on computer vision and image processing. I’m particularly interested in image restoration (e.g., real-world dehazing and depth estimation). Representative papers are highlighted.

NVDDE: Joint Non-aligned Regularization Video Dehazing and Depth Estimation in Real-world Scenes
Junkai Fan, Kun Wang, Zhiqiang Yan, Jun Li*, Jian Yang*

In review, 2024
project page / github

We propose a novel non-aligned learning framework to address video dehazing and depth estimation simultaneously by using the disentangled atmospheric scattering model.

Driving-Video Dehazing with Non-Aligned Regularization for Safety Assistance
Junkai Fan, Jiangwei Weng, Kun Wang, Yijun Yang, Jianjun Qian, Jun Li*, Jian Yang*

CVPR, 2024
project page / github / video

We present an innovative video dehazing framework for real-world driving scenarios, addressing temporal and spatial misalignment challenges with non-aligned hazy/clear video pairs and a reference frame matching module.

Non-aligned supervision for Real Image Dehazing
Junkai Fan, Fei Guo, Jianjun Qian, Xiang Li, Jun Li*, Jian Yang*

arXiv, 2023
project page / github

Given an image or video captured in a real foggy scene, our model is capable of restoring the corresponding clear scene image or video. Moreover, training our model does not require fully aligned ground truth (GT), which helps us collect real hazy scene data.

Selected Honors and Awards
  • 2018, The First Prize of Scholarship, Rank (4/38), Wenzhou University;
  • 2018, "Xiaoan Wang Award" for Innovation and Entrepreneurship, Rank (2/38), Wenzhou University;
  • 2017, The Graduate Scientific Research Foundation of Wenzhou University, Rank (1/12), Wenzhou University;

This webpage is fork from Jon Barron. Thanks to him!

Web Counters
Web Counters