├── .gitignore └── README.md /.gitignore: -------------------------------------------------------------------------------- 1 | *.pdf 2 | 3 | .DS_Store 4 | -------------------------------------------------------------------------------- /README.md: -------------------------------------------------------------------------------- 1 | # Deep Image Processing 2 | 3 | - Deep learning in image processing (especially real-world photography) 4 | 5 | - Similar repos: [deep-imaging](https://github.com/mdelbra/deep-imaging.git) & [Neural Networks for Low Level Image Processing](https://github.com/holovincent/Neural-Networks-for-Low-Level-Image-Processing) 6 | - Seems not maintained anymore 7 | 8 | *** 9 | 10 | ## Processing Pipeline 11 | 12 | - Transform Your Smartphone into a DSLR Camera: Learning the ISP in the Wild (ECCV 2022): [Paper](https://www.ecva.net/papers/eccv_2022/papers_ECCV/html/2134_ECCV_2022_paper.php) / [GitHub (code to be committed)](https://github.com/4rdhendu/TransformPhone2DSLR) 13 | 14 | - CameraNet: A Two-Stage Framework for Effective Camera ISP Learning (IEEE Transactions on Image Processing, 2020): [Paper](http://www4.comp.polyu.edu.hk/~cslzhang/paper/CameraNet.pdf) / [GitHub](https://github.com/ilegendforever/CameraNet_official) 15 | 16 | - Single-Image HDR Reconstruction by Learning to Reverse the Camera Pipeline (CVPR 2020): [Website](https://www.cmlab.csie.ntu.edu.tw/~yulunliu/SingleHDR) / [Paper](https://arxiv.org/abs/2004.01179) / [GitHub](https://github.com/alex04072000/SingleHDR) 17 | 18 | - Hardware-in-the-loop End-to-end Optimization of Camera Image Processing Pipelines: [Website](https://www.cs.princeton.edu/~fheide/HardwareInTheLoop-ImageOptimization/) / [Paper](https://www.cs.princeton.edu/~fheide/HardwareInTheLoop-ImageOptimization/images/HardwareInTheLoop_ImageOptimization.pdf) 19 | 20 | - Hyperparameter Optimization in Black-box Image Processing using Differentiable Proxies: [Website](https://www.cs.princeton.edu/~fheide/proxyopt) / [Paper](https://www.cs.princeton.edu/~fheide/ProxyOpt.pdf) 21 | 22 | - Replacing Mobile Camera ISP with a Single Deep Learning Model: [Website](http://www.vision.ee.ethz.ch/~ihnatova/pynet.html) / [Paper](https://arxiv.org/abs/2002.05509) / [GitHub](https://github.com/aiff22/PyNET) 23 | 24 | - Zoom to Learn, Learn to Zoom (CVPR 2019): [Website](https://people.eecs.berkeley.edu/~cecilia77/project-pages/zoom.html) / [Paper](https://arxiv.org/pdf/1905.05169.pdf) / [GitHub](https://github.com/ceciliavision/zoom-learn-zoom) 25 | 26 | - DeepISP: Towards Learning an End-to-End Image Processing Pipeline (IEEE TIP 2019): [Paper](https://arxiv.org/abs/1801.06724) / [Dataset](https://www.kaggle.com/knn165897/s7-isp-dataset) 27 | 28 | - VisionISP: Repurposing the Image Signal Processor for Computer Vision Applications (ICIP 2019): [IEEE](https://ieeexplore.ieee.org/document/8803607) 29 | 30 | - Deep Camera: A Fully Convolutional Neural Network for Image Signal Processing (ICCV 2019): [Paper](https://arxiv.org/ftp/arxiv/papers/1908/1908.09191.pdf) 31 | 32 | - CameraNet: A Two-Stage Framework for Effective Camera ISP Learning: [Paper](https://arxiv.org/abs/1908.01481) 33 | 34 | - DEEP LEARNING for Image and Video Processing: [Slide](http://home.ku.edu.tr/~mtekalp/Tutorial_slides.pdf) 35 | 36 | - Modelling the Scene Dependent Imaging in Cameras with a Deep Neural Network (ICCV 2017): [Paper](http://snam.ml/assets/publication/radiometricCal_iccv17/radiometricCal_iccv17.pdf) / [Supplementary Material](http://snam.ml/assets/publication/radiometricCal_iccv17/radiometricCal_iccv17_supp.pdf) 37 | 38 | - DSLR-Quality Photos on Mobile Devices with Deep Convolutional Networks (ICCV 2017): [Paper](https://arxiv.org/abs/1704.02470) / [Website](http://people.ee.ethz.ch/~ihnatova/) 39 | 40 | - Deep Semantics-Aware Photo Adjustment: [Paper](https://arxiv.org/abs/1706.08260) 41 | 42 | - Automatic Photo Adjustment Using Deep Neural Networks (ACM Transactions on Graphics, 2015): [Paper](https://arxiv.org/abs/1412.7725) / [GitHub](https://github.com/stephenyan1231/dl-image-enhance) / [Website](https://sites.google.com/site/homepagezhichengyan/home/dl_img_adjust) 43 | 44 | *** 45 | 46 | ## Exposure Correction 47 | 48 | - Learning Multi-Scale Photo Exposure Correction: [Paper](https://arxiv.org/abs/2003.11596) / [GitHub](https://github.com/mahmoudnafifi/Exposure_Correction) / [PyTorch version](https://github.com/LZ-CH/Exposure_Correction-pytorch) 49 | 50 | *** 51 | 52 | ## Image Enhancement 53 | 54 | - CURL: Neural Curve Layers for Image Enhancemen: [Paper](https://arxiv.org/pdf/1911.13175.pdf) / [GitHub](https://github.com/sjmoran/CURL) 55 | 56 | - Deep Photo Enhancer: Unpaired Learning for Image Enhancement from Photographs with GANs (CVPR 2018): [GitHub](https://github.com/nothinglo/Deep-Photo-Enhancer) 57 | 58 | *** 59 | 60 | ## Aesthetics 61 | 62 | - Similar GitHub repo: [Awesome Image Aesthetic Assessment and Cropping](https://github.com/bcmi/Awesome-Aesthetic-Evaluation-and-Cropping) 63 | 64 | - Composition and Style Attributes Guided Image Aesthetic Assessment: [Paper](https://arxiv.org/abs/2111.04647) 65 | 66 | - Neural Aesthetic Image Reviewer: [Paper](https://arxiv.org/abs/1802.10240) 67 | 68 | - A Deep Learning Perspective on Beauty, Sentiment, and Remembrance of Art: [Paper](http://fulir.irb.hr/4914/1/CetinicE_DeepLearningIEEEaccess2019_%207_73694.pdf) 69 | 70 | - Image Aesthetics Assessment using Fully Convolutional Neural Networks: [Paper](https://www.iti.gr/~bmezaris/publications/mmm19_lncs11295_1_preprint.pdf) 71 | 72 | - Neural Image Assessment: [Google AI Blog](https://ai.googleblog.com/2017/12/introducing-nima-neural-image-assessment.html) / [Paper](https://arxiv.org/abs/1709.05424) 73 | 74 | - Personalized Image Aesthetics: [Paper](https://openaccess.thecvf.com/content_iccv_2017/html/Ren_Personalized_Image_Aesthetics_ICCV_2017_paper.html) / [GitHub](https://github.com/alanspike/personalizedImageAesthetics) 75 | 76 | - RAPID: Rating Pictorial Aesthetics using Deep Learning: [Paper](http://infolab.stanford.edu/~wangz/project/imsearch/Aesthetics/ACMMM2014/lu.pdf) 77 | 78 | - [Papers with Code: Aesthetics Quality Assessment](https://paperswithcode.com/task/aesthetics-quality-assessment) 79 | 80 | *** 81 | 82 | ## Cropping/Composition 83 | 84 | - Composing Photos Like a Photographer (CVPR 2021): [Paper](https://openaccess.thecvf.com/content/CVPR2021/papers/Hong_Composing_Photos_Like_a_Photographer_CVPR_2021_paper.pdf) / [Supplement](https://openaccess.thecvf.com/content/CVPR2021/supplemental/Hong_Composing_Photos_Like_CVPR_2021_supplemental.pdf) 85 | 86 | - Reliable and Efficient Image Cropping: A Grid Anchor based Approach (CVPR 2019): [Paper](http://openaccess.thecvf.com/content_CVPR_2019/papers/Zeng_Reliable_and_Efficient_Image_Cropping_A_Grid_Anchor_Based_Approach_CVPR_2019_paper.pdf) / [GitHub1](https://github.com/HuiZeng/Grid-Anchor-based-Image-Cropping.git) / [GitHub2](https://github.com/HuiZeng/Grid-Anchor-based-Image-Cropping-Pytorch.git) / [GitHub3](https://github.com/lld533/Grid-Anchor-based-Image-Cropping-Pytorch.git) 87 | 88 | - Speedy Neural Networks for Smart Auto-Cropping of Images: [Website](https://blog.twitter.com/engineering/en_us/topics/infrastructure/2018/Smart-Auto-Cropping-of-Images.html) 89 | 90 | - [Papers with Code: Image Cropping](https://paperswithcode.com/task/image-cropping) 91 | 92 | - Similar repo: [Awesome-Aesthetic-Evaluation-and-Cropping](https://github.com/bcmi/Awesome-Aesthetic-Evaluation-and-Cropping) 93 | 94 | *** 95 | 96 | ## Demosaicing 97 | 98 | - Learning Deep Convolutional Networks for Demosaicing: [Website](https://www.cmlab.csie.ntu.edu.tw/project/Deep-Demosaic/) / [Paper](https://arxiv.org/abs/1802.03769) 99 | 100 | - Deep Demosaicing using ResNet-Bottleneck Architecture: [Paper](https://www.easychair.org/publications/preprint/NFnj) 101 | 102 | - Trinity of Pixel Enhancement: a Joint Solution for Demosaicking, Denoising and Super-Resolution: [Paper](https://arxiv.org/abs/1905.02538) / [GitHub](https://github.com/guochengqian/TENet.git) 103 | 104 | - Learning to Compose with Professional Photographs on the Web (ACM Multimedia 2017): 105 | [Paper](https://arxiv.org/abs/1702.00503) / [GitHub](https://github.com/yiling-chen/view-finding-network) / [GitHub](https://github.com/remorsecs/pytorch-view-finding-network) 106 | 107 | - Quantitative Analysis of Automatic Image Cropping Algorithms: A Dataset and Comparative Study (IEEE WACV 2017): [Website](https://yiling-chen.github.io/flickr-cropping-dataset/) / [Paper](http://arxiv.org/abs/1701.01480) / [GitHub](https://github.com/yiling-chen/flickr-cropping-dataset) 108 | 109 | - [Papers with Code: Demosaicking](https://paperswithcode.com/task/demosaicking) 110 | 111 | *** 112 | 113 | ## White Balance/Illuminant Estimation/Color Constancy 114 | 115 | - Object-based color constancy in a deep neural network (JOSA A 2023): [Paper](https://opg.optica.org/josaa/fulltext.cfm?uri=josaa-40-3-A48&id=525626) 116 | 117 | - A Multi-Hypothesis Approach to Color Constancy: [Paper](https://arxiv.org/pdf/2002.12896.pdf) / [GitHub](https://github.com/huawei-noah/multi_hyp_cc) 118 | 119 | - Deep White-Balance Editing: [Paper](https://arxiv.org/abs/2004.01354) / [GitHub](https://github.com/mahmoudnafifi/Deep_White_Balance) 120 | 121 | - Dual Illumination Estimation for Robust Exposure Correction: [Paper](https://arxiv.org/pdf/1910.13688.pdf) 122 | 123 | - Formulating Camera-Adaptive Color Constancy as a Few-shot Meta-Learning Problem: [Paper](https://arxiv.org/pdf/1811.11788.pdf) 124 | 125 | - Semantic White Balance: Semantic Color Constancy Using Convolutional Neural Network: [Paper](https://arxiv.org/pdf/1802.00153.pdf) 126 | 127 | - FC4: Fully Convolutional Color Constancy with Confidence-weighted Pooling (CVPR 2017): [Paper](http://openaccess.thecvf.com/content_cvpr_2017/papers/Hu_FC4_Fully_Convolutional_CVPR_2017_paper.pdf) / [GitHub](https://github.com/yuanming-hu/fc4) 128 | 129 | - [Papers with Code: Color Constancy](https://paperswithcode.com/task/color-constancy) 130 | 131 | - Related repo: [Awesome Color Constancy](https://github.com/iamsiddhantsahu/awesome-color-constancy.git) 132 | 133 | - [Color Constancy - Research Website on Illuminant Estimation](http://colorconstancy.com/) 134 | 135 | *** 136 | 137 | ## Denoising/Noise Modeling 138 | 139 | - CycleISP: Real Image Restoration via Improved Data Synthesis (CVPR 2020): [Paper](https://arxiv.org/abs/2003.07761) / [GitHub](https://github.com/swz30/CycleISP) 140 | 141 | - Generating Training Data for Denoising Real RGB Images 142 | via Camera Pipeline Simulation: [Website](https://people.csail.mit.edu/tiam/camera_sim/) 143 | 144 | - A Holistic Approach to Cross-Channel Image Noise Modeling and its Application to Image Denoising (CVPR 2016): [Paper](http://snam.ml/assets/publication/ccnoise_cvpr16/ccnoise_cvpr16.pdf) / [Supplementary Material](http://snam.ml/assets/publication/ccnoise_cvpr16/ccnoise_cvpr16_supp.pdf) / [Website](http://snam.ml/research/ccnoise/) / [GitHub](https://github.com/woozzu/ccnoise) 145 | 146 | *** 147 | 148 | ## Reversion (RGB -> RAW) 149 | 150 | - Learning sRGB-to-Raw-RGB De-rendering with Content-Aware Metadata (CVPR 2022): [Paper](https://openaccess.thecvf.com/content/CVPR2022/papers/Nam_Learning_sRGB-to-Raw-RGB_De-Rendering_With_Content-Aware_Metadata_CVPR_2022_paper.pdf) / [GitHub](https://github.com/SamsungLabs/content-aware-metadata) 151 | 152 | *** 153 | 154 | ## Datasets 155 | 156 | - MIT-Adobe FiveK Dataset: [Website](https://data.csail.mit.edu/graphics/fivek/) 157 | 158 | - Smartphone Image Denoising Dataset: [Website](https://www.eecs.yorku.ca/~kamel/sidd/index.php) 159 | --------------------------------------------------------------------------------