图像恢复raw2RGB
参考文章:
Google Drive资源下载（附DPED dataset百度云链接）, 谷歌云盘空间受限, 仅15G
https://blog.csdn.net/tuuzhang/article/details/86539507
AdaBN(Adaptive Batch Normalization)使用, CameraNet中使用
https://blog.csdn.net/qq_22764813/article/details/107044825
AndroidCamera开发
https://github.com/aserbao/AndroidCamera
MAI比赛, 同NITRE, 都是CVPR
CVF Open Access：https://openaccess.thecvf.com/menu
CVPR 2021 Open Access Repository：https://openaccess.thecvf.com/CVPR2021
ECVA | European Computer Vision Association：https://www.ecva.net/papers.php
查找CVPR、ECCV、ICCV文章链接
https://blog.csdn.net/qq_41542989/article/details/114885890
MAI 2021 Workshop, 包括raw2RGB(AWB/demosaic/gamma), raw降噪, HDR, 图像超分, 视频超分, 深度估计, camera场景检测
https://ai-benchmark.com/workshops/mai/2021/
MAI 2021 Workshop，模型测试
https://github.com/aiff22/MAI-2021-Workshop
https://ai-benchmark.com/tests.html
https://ai-benchmark.com/download
NTIRE2021: New Trends in Image Restoration and Enhancement workshop and challenges on image and video restoration and enhancement
https://data.vision.ee.ethz.ch/cvl/ntire21/
Team skyb solution for the AIM2020 mobile image signal processing challenge, 贡献了PyNet
https://github.com/egyptdj/skyb-aim2020-public
AIM2020: Advances in Image Manipulation workshop and challenges on image and video manipulation, 有raw2sRGB比赛, 感知性能恢复比赛
https://data.vision.ee.ethz.ch/cvl/aim20//
AIM2019: Advances in Image Manipulation workshop and challenges on image and video manipulation
https://data.vision.ee.ethz.ch/cvl/aim19/
raw2RGB数据集, dng格式raw图下载, P30专业版可保存dng文件
MIT-Adobe FiveK Dataset 图片自动下载
https://www.cnblogs.com/AlgrithmsRookie/p/11550223.html
MIT FiveK图像转化--DNG到TIFF，TIFF到JPEG
https://blog.csdn.net/LingTing00101/article/details/101855410
MIT fiveK官网, 数据集
https://data.csail.mit.edu/graphics/fivek/
The MIRFLICKR Retrieval Evaluation, 数据集
https://press.liacs.nl/mirflickr/
Preparing data for the MIT Adobe FiveK Dataset with Lightroom
HDR+ Burst Processing Pipeline, HDR+数据集
https://www.timothybrooks.com/tech/hdr-plus/
raw2sRGB+RGB图像增强网络
【论文详解】DPED：DSLR-Quality Photos on Mobile Devices with Deep Convolutional Networks
https://blog.csdn.net/lyl771857509/article/details/84405388
DPED, Software and pre-trained models for automatic photo quality enhancement using Deep Convolutional Networks
https://github.com/aiff22/DPED
DPED项目成果展示, DSLR-Quality Photos on Mobile Devices with Deep Convolutional Networks
http://people.ee.ethz.ch/~ihnatova/#dataset
PyTorch implement of RAW2RGBNet, codalab比赛成果展示和源码
https://github.com/MurrayC7/RAW2RGBNet
DeepISP, 非官方实现
https://github.com/MurrayC7/DeepISP
DeepISP: Learning End-to-End Image Processing Pipeline, 官方成果展示
https://elischwartz.github.io/DeepISP/
CameraNet_official, 官方暂未有代码公开
https://github.com/ilegendforever/CameraNet_official
CameraNet, 成果展示
https://github.com/ilegendforever/CameraNet
PyNet
Generating RGB photos from RAW image files with PyNET (PyTorch)
https://github.com/aiff22/PyNET-PyTorch
成果展示与数据集下载
http://people.ee.ethz.ch/~ihnatova/pynet.html
实现背景虚化效果
http://people.ee.ethz.ch/~ihnatova/pynet-bokeh.html
AWNet
AWNet-Attentive-Wavelet-Network-for-Image-ISP
https://github.com/Charlie0215/AWNet-Attentive-Wavelet-Network-for-Image-ISP
ISP基础_raw和RGB图像对获取
论文眼中的 isp流水线
https://blog.csdn.net/titer1/article/details/40866935
5分钟理解相机ISP（图像信号处理）
http://www.voycn.com/article/5fenzhonglijiexiangjiisptuxiangxinhaochuli
宅在家里涨知识：揭秘手机拍照原理，原来芯片里的ISP这么重要, 麒麟990官方宣传
https://baijiahao.baidu.com/s?id=1660383584242476773&wfr=spider&for=pc
佳能5D3逆天破解 无损RAW视频拍摄完整攻略, 魔灯官方
https://dcdv.zol.com.cn/377/3778167_all.html
视频素材中的RAW格式
https://zhuanlan.zhihu.com/p/154518913
关于RAW格式照片 你不可不知的十件事
https://baijiahao.baidu.com/s?id=1642244967796567828&wfr=spider&for=pc
为什么有的800w像素镜头可以拍1080p视频，有的1200w只能拍720p视频？
https://www.zhihu.com/question/21017257/answer/17659111
谷歌的秘密武器——Pixel Visual Core：需要仔细观察隐藏芯片, 谷歌类AI_ISP解决方案
https://baijiahao.baidu.com/s?id=1616543999280554129&wfr=spider&for=pc
怎么使用华为p30pro拍照的各种功能？个人用了两个月质量还是杠杠的？
http://www.360doc.com/content/20/0103/15/57193316_883916582.shtml
MOS人工图像质量评价方法, 亚马逊平台
Introduction to Amazon Mechanical Turk
https://docs.aws.amazon.com/AWSMechTurk/latest/AWSMechanicalTurkGettingStartedGuide/SvcIntro.html
Amazon Mechanical Turk 前景如何？
https://www.zhihu.com/question/30185115/answer/136697354
raw2sRGB损失函数, 还有VGG loss, 感知损失函数
PyTorch中的损失函数--L1Loss /L2Loss/SmoothL1Loss
https://zhuanlan.zhihu.com/p/83131026
Pytorch模型训练(4) - Loss Function, 主要为目标检测和目标分类
https://blog.csdn.net/u011681952/article/details/86607873
MS-SSIM loss, pytorch实现
https://github.com/VainF/pytorch-msssim
论文——《Loss Functions for Image Restoration With Neural Networks》, PL4NN_loss实现, caffe
https://blog.csdn.net/zyr_freedom/article/details/90381577
https://github.com/NVlabs/PL4NN/blob/master/src/loss.py
lightroom手动调色_raw和RGB图像对获取
色彩管理，详细剖析lightroom中HSL的使用方法
https://www.16xx8.com/photoshop/jiaocheng/2017/145115.html
HSL调色，高！大！上！
http://www.360doc.com/content/20/0421/08/16921388_907380612.shtml
简直“神技”：LR的HSL功能操作简单，能调所有色彩！
http://www.360doc.com/content/20/1022/21/6710993_941793651.shtml
rgb和Lab，rgb和hsl的色彩空间转换, python代码实现
https://blog.csdn.net/sunny_xsc1994/article/details/78541079
颜色空间RGB与HSV(HSL)的转换
https://blog.csdn.net/jiangxinyu/article/details/8000999
YCbCr色彩空间 （笔记1）
https://blog.csdn.net/qq_36955294/article/details/83443317
学习opencv之cvtColor
https://blog.csdn.net/weixin_42730667/article/details/102299280
其他
NIND论文jpg图像降噪, Master thesis on natural image noise removal using Convolutional Neural Networks, Natural Image Noise Dataset
https://github.com/trougnouf/nind-denoise
https://github.com/trougnouf/mthesis-denoise
深度学习添加水印，了解到相关论文
https://github.com/pkorus/neural-imaging
https://pkorus.pl/downloads/neural-imaging-resources
Lightweight & high-performance learned image compression codec (Tensorflow).深度学习图像压缩