Kangfu MEI's Homepage

Kangfu MEI

I am a Ph.D. student at Johns Hopkins University , advised by Prof. Vishal Patel , where I work on compute vision and computational photography. Before that, I was a M.Phil student at The Chinese University of Hong Kong, Shenzhen, advised by Prof. Rui Huang.

I previously interned at: DAMO Academy / Kuaishou / JD.COM

Email / CV / Google Scholar / Github / Zhihu

Research

My research interest mainly focuses on degraded images restoration as well as its applicatoin in high-level vision. Representative papers are highlighted.

	SDAN: Squared Deformable Alignment Network for Learning Misaligned Optical Zoom Kangfu Mei, Shenglong Ye, Rui Huang ICME, 2021 code / arXiv New state of the art of LearnZoom dataset. A squared deformable network is proposed for the misaligned image restoration learning.
	AttaNet: Attention-Augmented Network for Fast and Accurate Scene Parsing Qi Song, Kangfu Mei, Rui Huang AAAI, 2021 code / arXiv Two novel Strip Attention Module (SAM) and Attention Fusion Module (AFM) are proposed for enhancing the accuracy of semantic segmentation networks with limited computational complexity increasing. , espcically the scenes contains vertical strip areas
	Spatial Pixel Irrelevant Network for Learning Image-to-image Translation Kangfu Mei, Haoyu Wu, Juncheng Li, Rui Huang arXiv, 2020 Work in Progress Enable networks to learn the image translation without aligned pairs and achieve the state-of-the-art results in the real-world super-resolution task.
	Visual and Semantic Scene Understanding Under Bad Weather Kangfu Mei, Haoyu Wu, Juncheng Li, Rui Huang arXiv, 2019 Work in Progress Address the semantic segmentation accuracy dropping problem in dehazed images using the proposed semantic guiding loss.
	Higher-resolution Network for Image Demosaicing and Enhancing Kangfu Mei, Juncheng Li, Jiajie Zhang, Haoyu Wu, Jie Li, Rui Huang ICCV Worshop, 2019 (AIM2019 Winner Award) code / bibtex Achieved the best visual quality in Advances in Image Manipulation Workshop (AIM2019) Challenge on RAW to RGB Mapping.
	Lightweight and Accurate Recursive Fractal Network for Image Super-Resolution Juncheng Li, Yiting Yuan, Kangfu Mei, Faming Fang ICCV Worshop, 2019 (Oral Presentation) code / bibtex A lightweight Super-resolution framework that can construct infinitely possible topological sub-structure through Fractal Tree.
	Progressive Feature Fusion Network for Realistic Image Dehazing Kangfu Mei, Aiwen Jiang, Juncheng Li, Mingwen Wang ACCV, 2018 code / bibtex The first proposed haze removal method that removes haze from a single image in the complete end-to-end manner.
	Deep Residual Refining based Pseudo-multi-frame Network for Effective Single Image Super-resolution Kangfu Mei, Aiwen Jiang, Juncheng Li, Bo Liu, Jihua Ye, Mingwen Wang IET Image Processing, 2018 code / bibtex The first proposed haze removal method that removes haze from a single image in the complete end-to-end manner.
	Multi-scale Residual Network for Image Super-resolution Juncheng Li, Faming Fang, Kangfu Mei, Guixu Zhang ECCV, 2018 code / bibtex Introduce a novel multi-scale residual network for recovering the high-quality image from low-resolution.
	Effective Single-image Super-resolution Model using Squeeze-and-excitation Networks Kangfu Mei, Aiwen Jiang, Juncheng Li, Jihua Ye, Mingwen Wang ICNOIP, 2018 arXiv / code / bibtex Using squeeze-and-excitation blocks to improve the visual quality of super-resolution results with finer texture details.