michaelcai0912 / low_light_enhance_method

本仓库收集了目前有效的低光增强方法; This repositorycollects the low-light enhancement methods.

Introduction

low_light_enhance_method 是一个记录了近几年来优秀的低光增强算法的可以即插即用的仓库。

目前实现的算法

	算法名称
2022	URetinex_Net: URetinex-Net: Retinex-Based Deep Unfolding Network for Low-Light Image Enhancement
2022	SCI: Toward Fast, Flexible, and Robust Low-Light Image Enhancement
2022	SNR_LLIE_Net: SNR-aware Low-Light Image Enhancement
2020	SIM_CycleGAN: Lane Detection in Low-light Conditions Using an Efficient Data Enhancement : Light Conditions Style Transfer(没有预训练权重)
2021	RUAS: Retinex-inspired unrolling with cooperative prior architecture search for low-light image enhancement
2021	Zero-DCE++: Learning to enhance low-light image via zero-reference deep curve estimation
2021	EnlightenGAN: EnlightenGAN: Deep light enhancement without paired supervision
2020	Zero-DCE: Zero-reference deep curve estimation for low-light image enhancement

Getting Started

conda create -n low_light_enhance python=3.8
conda activate low_light_enhance
conda install pytorch==1.11.0 torchvision==0.12.0 torchaudio==0.11.0 cudatoolkit=11.3 -c pytorch
pip install opencv-python==4.6.0

真实低光图片上的增强效果

	增强前	增强后
URetinex-Net
SCI
SNR_LLIE_Net
RUAS
Zero-DCE++
Zero-DCE
EnlightenGAN