CVPR2020｜图像重建(超分辨率，图像恢复，去雨，去雾，去模糊，去噪等)相关论文汇总（附论文链接/开源代码/解析）【持续更新】

目录

整理了下今年CVPR图像重建相关的一些论文，包括超分辨率，图像恢复，去雨，去雾，去模糊，去噪等方向。大家如果觉得有帮助，欢迎点赞和收藏~~
Github地址如下，欢迎star~
https://github.com/Kobaayyy/Awesome-CVPR2020-Image-Reconstruction

CVPR2020的所有论文：http://openaccess.thecvf.com/CVPR2020.py
CVPR2020Workshops：http://openaccess.thecvf.com/CVPR2020_workshops/menu.py

1.超分辨率

图像超分辨率

PULSE: Self-Supervised Photo Upsampling via Latent Space Exploration of Generative Models

论文：https://arxiv.org/abs/2003.03808
代码：https://github.com/adamian98/pulse
解析：杜克大学提出 AI 算法，拯救渣画质马赛克秒变高清
备注：自监督；GAN；放大像素64倍（暂时是最高倍数）；将生成HR图像对应的LR图像与原图（LR）对比，找到最接近的那张，并反推找到对应的HR图像

Closed-Loop Matters: Dual Regression Networks for Single Image Super-Resolution

论文：https://arxiv.org/abs/2003.07018
代码：https://github.com/guoyongcs/DRN
解析：
备注：DRN

EventSR: From Asynchronous Events to Image Reconstruction, Restoration, and Super-Resolution via End-to-End Adversarial Learning

作者： Lin Wang, Tae-Kyun Kim, Kuk-Jin Yoon
单位：韩国科学技术院；伦敦帝国学院
论文：http://openaccess.thecvf.com/content_CVPR_2020/papers/Wang_EventSR_From_Asynchronous_Events_to_Image_Reconstruction_Restoration_and_Super-Resolution_CVPR_2020_paper.pdf
视频 ：https://www.youtube.com/watch?v=OShS_MwHecs
数据集： https://github.com/wl082013/ESIM_dataset
备注：图像重建、恢复、超分

Unpaired Image Super-Resolution Using Pseudo-Supervision

论文：https://arxiv.org/abs/2002.11397?context=eess
代码：
解析：#每日五分钟一读#Image Super-Resolution
备注：

Correction Filter for Single Image Super-Resolution: Robustifying Off-the-Shelf Deep Super-Resolvers

作者： Shady Abu Hussein, Tom Tirer, Raja Giryes
论文：https://arxiv.org/abs/1912.00157

Residual Feature Aggregation Network for Image Super-Resolution

论文：http://openaccess.thecvf.com/content_CVPR_2020/papers/Liu_Residual_Feature_Aggregation_Network_for_Image_Super-Resolution_CVPR_2020_paper.pdf
代码：
解析：超越RCAN，图像超分又一峰：RFANet
备注：超越RCAN，图像超分又一峰：RFANet

Deep Unfolding Network for Image Super-Resolution

论文：https://arxiv.org/abs/2003.10428
代码：https://github.com/cszn/USRNet
解析：CVPR2020：USRNet
备注：USRNet

Image Super-Resolution With Cross-Scale Non-Local Attention and Exhaustive Self-Exemplars Mining

论文：https://arxiv.org/abs/2006.01424
代码：https://github.com/SHI-Labs/Cross-Scale-Non-Local-Attention

Learning Texture Transformer Network for Image Super-Resolution

论文：https://arxiv.org/abs/2006.04139
代码：https://github.com/FuzhiYang/TTSR
备注：注意力机制

Robust Reference-Based Super-Resolution With Similarity-Aware Deformable Convolution

论文：http://openaccess.thecvf.com/content_CVPR_2020/html/Shim_Robust_Reference-Based_Super-Resolution_With_Similarity-Aware_Deformable_Convolution_CVPR_2020_paper.html

Structure-Preserving Super Resolution with Gradient Guidance

论文：https://arxiv.org/abs/2003.13063
代码：https://github.com/Maclory/Deep-Iterative-Collaboration
解析：CVPR2020丨SPSR：基于梯度指导的结构保留超分辨率方法
备注：SPSR

Unified Dynamic Convolutional Network for Super-Resolution With Variational Degradations

论文：https://arxiv.org/abs/2004.06965
代码：
解析：UDVD：适用于可变降质类型的通用图像超分，附参考代码
备注：UDVD

Perceptual Extreme Super Resolution Network with Receptive Field Block

论文：https://arxiv.org/abs/2005.12597
代码：
解析：NTIRE2020冠军方案RFB-ESRGAN：带感受野模块的超分网络
备注：NTIRE2020极限超分冠军方案RFB-ESRGAN；Workshops

Real-World Super-Resolution via Kernel Estimation and Noise Injection

论文：http://openaccess.thecvf.com/content_CVPRW_2020/html/w31/Ji_Real-World_Super-Resolution_via_Kernel_Estimation_and_Noise_Injection_CVPRW_2020_paper.html
代码：https://github.com/jixiaozhong/RealSR
解析：
备注：NTIRE2020-RWSR超分双赛道冠军方案；Workshops

Investigating Loss Functions for Extreme Super-Resolution

论文：http://openaccess.thecvf.com/content_CVPRW_2020/papers/w31/Jo_Investigating_Loss_Functions_for_Extreme_Super-Resolution_CVPRW_2020_paper.pdf
代码：https://github.com/kingsj0405/ciplab-NTIRE-2020
解析：
备注：NTIRE2020极限超分亚军方案CIPLab；Workshops

Nested Scale-Editing for Conditional Image Synthesis

论文：http://arxiv.org/abs/2006.02038
备注：解耦表征、多模图像转换、超分、修复

MSG-GAN: Multi-Scale Gradients for Generative Adversarial Networks

论文：https://arxiv.org/abs/1903.06048v3
代码：https://github.com/akanimax/msg-stylegan-tf
解析：CVPR2020之MSG-GAN：简单有效的SOTA
备注：NTIRE2020极限超分亚军方案CIPLab；Workshops

Guided Frequency Separation Network for Real-World Super-Resolution

论文：http://openaccess.thecvf.com/content_CVPRW_2020/papers/w31/Zhou_Guided_Frequency_Separation_Network_for_Real-World_Super-Resolution_CVPRW_2020_paper.pdf
代码：
解析：CVPR2020 ｜ 高低频分离超分方案
备注：NTIRE2020极限超分前五方案；Workshops

视频超分辨率

TDAN: Temporally-Deformable Alignment Network for Video Super-Resolution

论文：https://arxiv.org/abs/1812.02898
代码：https://github.com/YapengTian/TDAN-VSR-CVPR-2020
Demo Video：https://www.youtube.com/watch?v=eZExENE50I0
备注：首次将形变卷积用到视频超分领域；TDAN

Zooming Slow-Mo: Fast and Accurate One-Stage Space-Time Video Super-Resolution

论文：https://arxiv.org/abs/2002.11616
代码：https://github.com/Mukosame/Zooming-Slow-Mo-CVPR-2020
解析：慢镜头变焦：视频超分辨率：CVPR2020论文解析

Video Super-Resolution With Temporal Group Attention

论文：http://openaccess.thecvf.com/content_CVPR_2020/papers/Isobe_Video_Super-Resolution_With_Temporal_Group_Attention_CVPR_2020_paper.pdf

Space-Time-Aware Multi-Resolution Video Enhancement

主页：https://alterzero.github.io/projects/STAR.html
论文：http://arxiv.org/abs/2003.13170
代码：https://github.com/alterzero/STARnet

人脸超分辨率

Learning to Have an Ear for Face Super-Resolution

论文：https://arxiv.org/abs/1909.12780

Deep Face Super-Resolution With Iterative Collaboration Between Attentive Recovery and Landmark Estimation

论文：https://arxiv.org/abs/1812.02898
代码：https://github.com/YapengTian/TDAN-VSR-CVPR-2020

深度图超分辨率

Channel Attention Based Iterative Residual Learning for Depth Map Super-Resolution

论文：https://arxiv.org/abs/2006.01469

光场图像超分辨率

Light Field Spatial Super-Resolution via Deep Combinatorial Geometry Embedding and Structural Consistency Regularization

论文：https://arxiv.org/abs/2004.02215
代码：https://github.com/jingjin25/LFSSR-ATO

高光谱图像超分辨率

Unsupervised Adaptation Learning for Hyperspectral Imagery Super-Resolution

论文：http://openaccess.thecvf.com/content_CVPR_2020/papers/Zhang_Unsupervised_Adaptation_Learning_for_Hyperspectral_Imagery_Super-Resolution_CVPR_2020_paper.pdf
代码：https://github.com/JiangtaoNie/UAL

零样本超分辨率

Meta-Transfer Learning for Zero-Shot Super-Resolution

论文：https://arxiv.org/abs/2002.12213
代码：https://github.com/JWSoh/MZSR

用于超分辨率的数据增广

Rethinking Data Augmentation for Image Super-resolution: A Comprehensive Analysis and a New Strategy

论文：http://openaccess.thecvf.com/content_CVPR_2020/html/Yoo_Rethinking_Data_Augmentation_for_Image_Super-resolution_A_Comprehensive_Analysis_and_CVPR_2020_paper.html
代码：https://github.com/clovaai/cutblur

超分辨率用于语义分割

Dual Super-Resolution Learning for Semantic Segmentation

论文：http://openaccess.thecvf.com/content_CVPR_2020/html/Wang_Dual_Super-Resolution_Learning_for_Semantic_Segmentation_CVPR_2020_paper.html
代码：https://github.com/wanglixilinx/DSRL

2.图像恢复

Learning Invariant Representation for Unsupervised Image Restoration

论文：https://arxiv.org/pdf/2003.12769.pdf
代码：https://github.com/Wenchao-Du/LIR-for-Unsupervised-IR

Contextual Residual Aggregation for Ultra High-Resolution Image Inpainting

论文：https://arxiv.org/abs/2005.09704
备注：超高分辨率图像修复、注意力机制

UCTGAN: Diverse Image Inpainting based on Unsupervised Cross-Space

论文：http://openaccess.thecvf.com/content_CVPR_2020/papers/Zhao_UCTGAN_Diverse_Image_Inpainting_Based_on_Unsupervised_Cross-Space_Translation_CVPR_2020_paper.pdf
备注：图像修复、注意力机制

Attentive Normalization for Conditional Image Generation

论文：https://arxiv.org/abs/2004.03828
备注：注意力机制、类条件图像生成、图像修复

3.去雨

Deep Adversarial Decomposition: A Unified Framework for Separating Superimposed Images

论文：http://openaccess.thecvf.com/content_CVPR_2020/html/Zou_Deep_Adversarial_Decomposition_A_Unified_Framework_for_Separating_Superimposed_Images_CVPR_2020_paper.html

Multi-Scale Progressive Fusion Network for Single Image Deraining

论文：https://arxiv.org/abs/2003.10985

代码：https://github.com/kuihua/MSPFN

4.去雾

Domain Adaptation for Image Dehazing

论文：https://arxiv.org/abs/2005.04668

Multi-Scale Boosted Dehazing Network with Dense Feature Fusion

论文：https://arxiv.org/abs/2004.13388
代码：https://github.com/BookerDeWitt/MSBDN-DFF

5.去模糊

视频去模糊

Cascaded Deep Video Deblurring Using Temporal Sharpness Prior

主页：https://csbhr.github.io/projects/cdvd-tsp/index.html
论文：https://arxiv.org/abs/2004.02501
代码：https://github.com/csbhr/CDVD-TSP

6.去噪

A Physics-based Noise Formation Model for Extreme Low-light Raw Denoising

论文：https://arxiv.org/abs/2003.12751

代码：https://github.com/Vandermode/NoiseModel

CycleISP: Real Image Restoration via Improved Data Synthesis

论文：https://arxiv.org/abs/2003.07761

代码：https://github.com/swz30/CycleISP

未完待续~

参考

[1] 杜克大学提出 AI 算法，拯救渣画质马赛克秒变高清
[2] CVPR 2020 论文大盘点-超分辨率篇
[3] CVPR2020丨SPSR：基于梯度指导的结构保留超分辨率方法
[4] CVPR2020：USRNet
[5] UDVD：适用于可变降质类型的通用图像超分，附参考代码
[6] NTIRE2020冠军方案RFB-ESRGAN：带感受野模块的超分网络
[7] 超越RCAN，图像超分又一峰：RFANet
[8] #每日五分钟一读#Image Super-Resolution
[9] CVPR 2020 | 几篇GAN在low-level vision中的应用论文
[10] 超100篇！CVPR 2020最全GAN论文梳理汇总！
[11] CVPR2020之MSG-GAN：简单有效的SOTA
[12] CVPR2020-Code
[13] 慢镜头变焦：视频超分辨率：CVPR2020论文解析
[14] CVPR2020 ｜ 高低频分离超分方案
码字不易，如果您觉得有帮助，欢迎点赞和收藏~~

【推广】 免费学中医，健康全家人
原文地址：https://www.cnblogs.com/Kobaayyy/p/13163056.html