awesome-computer-vision-papers

A list of papers and other resources on computer vision and deep learning.

Reviews

• A Survey on Deep Learning Techniques for Stereo-based Depth Estimation. arXiv202006
• Deep Learning for LiDAR Point Clouds in Autonomous Driving: A Review. arXiv202005
• A Gentle Introduction to Deep Learning for Graphs. arXiv201912 [Note]
• A Comprehensive Survey on Graph Neural Networks. arXiv201912 [Note]
• Research Guide: Model Distillation Techniques for Deep Learning, Derrick Mwiti, 2019.11 [Blog]
• Graph Neural Networks: A Review of Methods and Applications, arXiv2019.7 [Intro-Chinese]
• A Review on Deep Learning in Medical Image Reconstruction, arXiv2019.6
• MNIST-C: A Robustness Benchmark for Computer Vision, arXiv2019.6 [Code&Dataset]
• Going Deep in Medical Image Analysis: Concepts, Methods, Challenges and Future Directions， arXiv2019.2
• Computer Vision for Autonomous Vehicles: Problems, Datasets and State-of-the-Art. arXiv201704 [Resourses]
• [2019TIV] A Survey of Autonomous Driving: Common Practices and Emerging Technologies
• [2014JMLR] Do we need hundreds of classifiers to solve real world classification problems

SemanticSeg

awesome-semantic-segmentation

SemanticSegPaperCollection

SegLoss: A collection of loss functions for medical image segmentation

Efficient-Segmentation-Networks

U-Net and its variant code

深度学习下的语义分割综述 [Page] [Notes]

三维语义分割概述及总结 [Page]

Unpooling/unsampling deconvolution [Note]

Some basic points: align_corners

Code: Semantic Segmentation Suite in TensorFlow

Review

• A Survey on Instance Segmentation: State of the art, arXiv202007
• Unsupervised Domain Adaptation in Semantic Segmentation: a Review, arXiv202005
• Image Segmentation Using Deep Learning: A Survey. arXiv202001
• Recent progress in semantic image segmentation, Artificial Intelligence Review, 2019
• Review of Deep Learning Algorithms for Image Semantic Segmentation, 2018 [Blog]

arXiv

• Divided We Stand: A Novel Residual Group Attention Mechanism for Medical Image Segmentation, arXiv2019.12
• Hard Pixels Mining: Learning Using Privileged Information for Semantic Segmentation, arXiv2019.11
• Hierarchical Attention Networks for Medical Image Segmentation, arXiv2019.11 [eye line seg]
• Multi-scale guided attention for medical image segmentation, arXiv2019.10 [Code]
• Adaptive Class Weight based Dual Focal Loss for Improved Semantic Segmentation, arXiv2019.10
• ELKPPNet: An Edge-aware Neural Network with Large Kernel Pyramid Pooling for Learning Discriminative Features in Semantic Segmentation, arXiv2019.6
• ESNet: An Efficient Symmetric Network for Real-time Semantic Segmentation, arXiv2019.6 [Code]
• FastFCN: Rethinking Dilated Convolution in the Backbone for Semantic Segmentation, arXiv2019.3 [Proj] [Code] [Note] [JPU: Joint Pyramid Upsampling]
• ENet: A Deep Neural Network Architecture for Real-Time Semantic Segmentation, arXiv2016.6 [Code]

Journal/Proceedings

• [2019IJCV] AdapNet++: Self-Supervised Model Adaptation for Multimodal Semantic Segmentation [Code]

• [2019NIPS] Zero-Shot Semantic Segmentation [Code]

• [2019NIPS] Grid Saliency for Context Explanations of Semantic Segmentation [github]

• [2019NIPS] Region Mutual Information Loss for Semantic Segmentation

• [2019NIPS] Improving Semantic Segmentation via Dilated Affinity

• [2019NIPS] Correlation Maximized Structural Similarity Lossfor Semantic Segmentation

• [2019NIPS] Multi-source Domain Adaptation for Semantic Segmentation

• [2019ICCV] Boundary-Aware Feature Propagation for Scene Segmentation

• [2019ICCV] [Adaptive-sampling] Efficient Segmentation: Learning Downsampling Near Semantic Boundaries [github] (Reference: LIP: Local Importance-based Pooling, ICCV2019 [github] [Notes])

• [2019ICCV] Selectivity or Invariance: Boundary-aware Salient Object Detection [Proj&Code]

• [2019ICCV] Recurrent U-Net for Resource-Constrained Segmentation

• [2019ICCV] Gated-SCNN: Gated Shape CNNs for Semantic Segmentation [Code] [Proj]

• [2019ICCV] Visualizing the Invisible: Occluded Vehicle Segmentation and Recovery

• [2019ICCV] ACE: Adapting to Changing Environments for Semantic Segmentation

• [2019ICCV] Asymmetric Non-local Neural Networks for Semantic Segmentation

• [2019ICCV] DADA: Depth-Aware Domain Adaptation in Semantic Segmentation

• [2019ICCV] ACFNet: Attentional Class Feature Network for Semantic Segmentation

• [2019ICCV] [EMANet] Expectation-Maximization Attention Networks for Semantic Segmentation [github]

• [2019ICCV] CCNet : Criss-Cross Attention for Semantic Segmentation [github]

• [2019ICCV] Gated-SCNN: Gated Shape CNNs for Semantic Segmentation

• [2019CVPR] ESPNetv2: A Light-weight, Power Efficient, and General Purpose Convolutional Neural Network [Code]

• [2019CVPR] Not All Areas Are Equal: Transfer Learning for Semantic Segmentation via Hierarchical Region Selection

• [2019CVPR] Beyond Gradient Descent for Regularized Segmentation Losses [Code]

• [2019CVPR] Co-occurrent Features in Semantic Segmentation

• [2019CVPR] Context-aware Spatio-recurrent Curvilinear Structure Segmentation [line structure seg]

• [2019CVPR] Dual attention network for scene segmentation

• [2019CVPR] Auto-DeepLab: Hierarchical Neural Architecture Search for Semantic Image Segmentation.

• [2019AAAI] Learning Fully Dense Neural Networks for Image Semantic Segmentation

• [2019MICCAI] ET-Net: A Generic Edge-Attention Guidance Network for Medical Image Segmentation [Code]

• [2019MICCAI] Attention Guided Network for Retinal Image Segmentation [Code]

• [2019MICCAIW] CU-Net: Cascaded U-Net with Loss Weighted Sampling for Brain Tumor Segmentation

• [2018CVPR] [EncNet] Context Encoding for Semantic Segmentation (oral) [Code-Pytorch] [Slides]

• [2018CVPR] Learning a Discriminative Feature Network for Semantic Segmentation

• [2018CVPR] DenseASPP for Semantic Segmentation in Street Scenes [Code]

• [2018CVPR] Dense Decoder Shortcut Connections for Single-Pass Semantic Segmentation

• [2018ECCV] ESPNet: Efficient Spatial Pyramid of Dilated Convolutions for Semantic Segmentation

• [2018ECCV] ICNet for Real-Time Semantic Segmentation on High-Resolution Images [Proj] [Code]

• [2018ECCV] PSANet: Point-wise Spatial Attention Network for Scene Parsing

• [2018ECCV] Bisenet: Bilateral segmentation network for real-time semantic segmentation [Code]

• [2018ECCV] [DeepLabv3+] Encoder-Decoder with Atrous Separable Convolution for Semantic Image Segmentation [Code]

• [2018BMVC] Pyramid Attention Network for Semantic Segmentation

• [2018DLMIA] UNet++: A Nested U-Net Architecture for Medical Image Segmentation [Code]

• [2018MIDL] Attention U-Net: Learning Where to Look for the Pancreas

• [2017arXiv] [DeepLabv3] Rethinking Atrous Convolution for Semantic Image Segmentation

• [2017PAMI] [DeepLabv2] DeepLab: Semantic Image Segmentation with Deep Convolutional Nets, Atrous Convolution, and Fully Connected CRFs

• [2017PAMI] SegNet: A deep convolutional encoder-decoder architecture for image segmentation

• [2017CVPR] [GCN] Large Kernel Matters-Improve Semantic Segmentation by Global Convolutional Network [Code] [Note]

• [2017CVPR] [PSPNet] Pyramid Scene Parsing Network

• [2017CVPR] RefineNet: Multi-path refinement networks for high-resolution semantic segmentation

• [2017CVPR] [FCIS] Fully convolutional instance-aware semantic segmentation

• [2017CVPR] [FRRN] Full-Resolution Residual Networks for Semantic Segmentation in Street Scenes [Code]

• [2017CVPRW] The One Hundred Layers Tiramisu: Fully Convolutional DenseNets for Semantic Segmentation [Code]

• [2017ICRA] AdapNet: Adaptive semantic segmentation in adverse environmental conditions [Code]

• [2016ICLR] Multi-Scale Context Aggregation by Dilated Convolutions

• [2016ICLR] ParseNet: Looking Wider to See Better

• [2016CVPR] Instance-aware semantic segmentation via multi-task network cascades

• [2016CVPR] Attention to Scale: Scale-Aware Semantic Image Segmentation

• [2016ECCV] What's the Point: Semantic Segmentation with Point Supervision

• [2016ECCV] Instance-sensitive fully convolutional networks

• [2016DLMIA] [UNet+ResNet] The Importance of Skip Connections in Biomedical Image Segmentation

• [2015ICLR] [DeepLabv1] Semantic Image Segmentation with Deep Convolutional Nets and Fully Connected CRFs

• [2015ICCV] Conditional random fields as recurrent neural networks

• [2015ICCV] [DeconvNet] Learning Deconvolution Network for Semantic Segmentation

• [2015MICCAI] U-Net: Convolutional networks for biomedical image segmentation [Note]

• [2015CVPR/2017PAMI] [FCN] Fully convolutional networks for semantic segmentation

PanopticSeg

• awesome-panoptic-segmentation

• Real-Time Panoptic Segmentation from Dense Detections, arXiv2019.12

• Panoptic-DeepLab: A Simple, Strong, and Fast Baseline for Bottom-Up Panoptic Segmentation, arXiv2019.12

• PanDA: Panoptic Data Augmentation, arXiv2019.11

• Learning Instance Occlusion for Panoptic Segmentation, arXiv2019.11

• Panoptic Edge Detection, arXiv2019.6

• [2020ICRA] DS-PASS: Detail-Sensitive Panoramic Annular Semantic Segmentation through SwaftNet for Surrounding Sensing [Code]

• [2020AAAI] SOGNet: Scene Overlap Graph Network for Panoptic Segmentation

• [2019CVPR] Panoptic Segmentation

• [2019CVPR] Attention-guided Unified Network for Panoptic Segmentation

• [2019CVPR] Panoptic Feature Pyramid Networks (oral) [unofficial code] [detectron2]

• [2019CVPR] UPSNet: A Unified Panoptic Segmentation Network [Code]

• [2019CVPR] [OANet] An End-to-end Network for Panoptic Segmentation

• [2019CVPR] DeeperLab: Single-Shot Image Parser (oral) [project] [code]

• [2019CVPR] Interactive Full Image Segmentation by Considering All Regions Jointly

• [2019CVPR] Seamless Scene Segmentation [code]

3DVision

awesome image-based 3D reconstruction

awesome-point-cloud-analysis

[Blog] 基于单目视觉的三维重建算法综述

[Bolg] 三维视觉、SLAM方向全球顶尖实验室汇总

Camera Calibration [Note] [Note2]

[Hub] Visual SLAM Related Research

Review

• BigSFM: Reconstructing the World from Internet Photos, summary of Noah Snavely works [Proj&Code] (Bundler, 1DSfM, sfm-dismbig, DISCO, LocalSymmetry, dataset ...)

• A Survey on Deep Leaning Architectures for Image-based Depth Reconstruction, arXiv2019.6

• [2019PAMI] Image-based 3D Object Reconstruction: State-of-the-Art and Trends in the Deep Learning Era

• [2017Robot] Keyframe-based monocular SLAM: design, survey, and future directions, Robotics and Autonomous Systems

CameraPose

• [2017CVPR] Geometric loss functions for camera pose regression with deep learning [Proj-with PoseNet+Modelling]
• [2016ICRA] Modelling Uncertainty in Deep Learning for Camera Relocalization
• [2015ICCV] PoseNet: A Convolutional Network for Real-Time 6-DOF Camera Relocalization

SfM

• [2016ECCV] [LineSfM] Robust and Accurate Line- and/or Point-Based Pose Estimation without Manhattan Assumptions [Code]

Depth/StereoMatching

• [2020PAMI] SurfaceNet+: An End-to-end 3D Neural Network for Very Sparse Multi-view Stereopsis [Code]
• [2020ICLR] Pseudo-LiDAR++: Accurate Depth for 3D Object Detection in Autonomous Driving, arXiv2019.8 [Code]
• [2019NIPS] [SC-SfMLearner] Unsupervised Scale-consistent Depth and Ego-motion Learning from Monocular Video [Proj] [Code]
• [2019ICCV] How do neural networks see depth in single images? [Note]
• [2019ICCV] DeepPruner: Learning Efficient Stereo Matching via Differentiable PatchMatch [Code]
• [2019CVPR] Pseudo-LiDAR from Visual Depth Estimation: Bridging the Gap in 3D Object Detection for Autonomous Driving [Code]
• [2019CVPR] DeepLiDAR: Deep Surface Normal Guided Depth Prediction for Outdoor Scene from Sparse LiDAR Data and Single Color Image
• [2019CVPR] [R-MVSNet] Recurrent MVSNet for High-resolution Multi-view Stereo Depth Inference [Code]
• [2019ToG] 3D Ken Burns Effect from a Single Image [Homepage] [Code]
• [2019IROS] SuMa++: Efficient LiDAR-based Semantic SLAM [Code]
• [2019ICCVW] Self-Supervised Learning of Depth and Motion Under Photometric Inconsistency
• [2019WACV] SfMLearner++: Learning Monocular Depth & Ego-Motion using Meaningful Geometric Constraints
• [2018CVPR] Automatic 3D Indoor Scene Modeling From Single Panorama
• [2018CVPR] LEGO: Learning Edge with Geometry all at Once by Watching Videos (spotlight) [Code]
• [2018CVPR] [vid2depth] Unsupervised Learning of Depth and Ego-Motion from Monocular Video Using 3D Geometric Constraints [Proj&Code]
• [2018CVPR] GeoNet: Unsupervised Learning of Dense Depth, Optical Flow and Camera Pose [Code]
• [2018CVPR] DeepMVS: Learning Multi-View Stereopsis [Proj] [Code]
• [2018ECCV] MVSNet: Depth Inference for Unstructured Multi-view Stereo
• [2017ICCV] SurfaceNet: An End-to-end 3D Neural Network for Multiview Stereopsis [Code]
• [2017CVPR] [SfMLearner] Unsupervised Learning of Depth and Ego-Motion from Video, Oral [Proj] [TF] [Pytorch] [ClassProj]
• [2017CVPR] SGM-Nets: Semi-Global Matching With Neural Networks
• [2016JMLR] [MC-CNN] Stereo matching by training a convolutional neural network to compare image patches [Code]

Surface Reconstruction

• [ICCV15/IJCV17] Global, Dense Multiscale Reconstruction for a Billion Points [Proj] [Code]
• [2014ECCV] Let there be color! Large-scale texturing of 3D reconstructions [Code]

3D Layout

• [2017WACV] Pano2CAD: Room Layout From A Single Panorama Image

• [2014ECCV] PanoContext: A Whole-room 3D Context Model for Panoramic Scene Understanding, Oral [Homepage&Code] [PanoBasic]

3D SemanticSeg

• Kimera: an Open-Source Library for Real-Time Metric-Semantic Localization and Mapping, arXiv2019.12 [Code]

• Rotation Invariant Point Cloud Classification: Where Local Geometry Meets Global Topology, arXiv2019.11

• SalsaNet: Fast Road and Vehicle Segmentation in LiDAR Point Clouds for Autonomous Driving, arXiv2019.9 [Code]

• Going Deeper with Point Networks, arXiv2019.7 [Code]

• [2020GRSM] A Review of Point Cloud Semantic Segmentation

• [2019NIPS] [PVCNN] Point-Voxel CNN for Efficient 3D Deep Learning (Spotlight) [Proj] [Code]

• [2019IROS] RangeNet++: Fast and Accurate LiDAR Semantic Segmentation [Code]

• [2019ICCV] SemanticKITTI: A Dataset for Semantic Scene Understanding of LiDAR Sequences

• [2019ICCV] Hierarchical Point-Edge Interaction Network for Point Cloud Semantic Segmentation

• [2019ICCV] Cascaded Context Pyramid for Full-Resolution 3D Semantic Scene Completion (oral)

• [2019CVPR] ClusterNet: Deep Hierarchical Cluster Network With Rigorously Rotation-Invariant Representation for Point Cloud Analysis

• [2018NIPS] PointCNN: Convolution On X-Transformed Points [Code]

• [2018ECCV] Efficient Semantic Scene Completion Network with Spatial Group Convolution [Code]

• [2017NIPS] PointNet++: Deep Hierarchical Feature Learning on Point Sets in a Metric Space [Code]

• [2017CVPR] PointNet: Deep Learning on Point Sets for 3D Classification and Segmentation [Code]

LowLevelVision

Tutorial&Reviews

• ICCV2019 Tutorial: Understanding Color and the In-Camera Image Processing Pipeline for Computer Vision, Michael S. Brown [Homepage] [Slides]
• CVPR2016 Tutorial: Understanding the In-Camera Image Processing Pipeline for Computer Vision, Michael S. Brown [Slides]
• NIPS2011 Tutorial: Modeling the Digital Camera Pipeline: From RAW to sRGB and Back, Michael S Brown [Slides]

RAW

• [2018IJCV] RAW Image Reconstruction Using a Self-contained sRGB–JPEG Image with Small Memory Overhead [Michael S. Brown]
• [2016CVPR] RAW Image Reconstruction using a Self-Contained sRGB-JPEG Image with only 64 KB Overhead
• [2014CVPR] Raw-to-raw: Mapping between image sensor color responses

Super-Resolution

• [Blog] [深入浅出深度学习超分辨率](https://mp.weixin.qq.com/s/o-I6T8f4AcETJqlDNZs9ug

• A Deep Journey into Super-resolution: A survey, arXiv2019.9

• [2020PAMI] Deep Learning for Image Super-resolution: A Survey

• [2019IJAC] Deep Learning Based Single Image Super-resolution: A Survey

• Densely Residual Laplacian Super-resolution, arXiv2019.7 [Code]

• Lightweight Image Super-Resolution with Adaptive Weighted Learning Network, arXiv2019.4 [Code]

• [2019SIGG] Handheld Multi-Frame Super-Resolution

• [2019CVPR] Deep Plug-and-Play Super-Resolution for Arbitrary Blur Kernels

• [2019CVPR] Zoom To Learn, Learn To Zoom [ProjPage] [Code]

• [2019CVPR] Towards Real Scene Super-Resolution with Raw Images [Code]

• [2019CVPR] 3D Appearance Super-Resolution with Deep Learning [Code]

• [2019CVPR] Learning Parallax Attention for Stereo Image Super-Resolution [Code]

• [2019CVPR] Meta-SR: A Magnification-Arbitrary Network for Super-Resolution [github]

• [2019CVPRW] Hierarchical Back Projection Network for Image Super-Resolution [Code]

• [2019ICCVW] Edge-Informed Single Image Super-Resolution [Code]

• [2017CVPRW] Enhanced Deep Residual Networks for Single Image Super-Resolution [Code]

• [2016PAMI] [SRCNN] Image Super-Resolution Using Deep Convolutional Networks

• [2016NIPS] Image Restoration Using Very Deep Convolutional Encoder-Decoder Networks with Symmetric Skip Connections

• [2016CVPR] [ESPCN] Real-Time Single Image and Video Super-Resolution Using an Efficient Sub-Pixel Convolutional Neural Network

• [2016CVPR] [VDSR] Accurate Image Super-Resolution Using Very Deep Convolutional Networks

• [2016ECCV] [FSRCNN] Accelerating the Super-Resolution Convolutional Neural Network

• [2014ECCV] [SRCNN] Learning a Deep Convolutional Network for Image Super-Resolution

Enhancement

• Diving Deeper into Underwater Image Enhancement: A Survey, arXiv2019.7
• [2018CVPR] Deep Photo Enhancer: Unpaired Learning for Image Enhancement from Photographs with GANs, [Homepage] [Code]
• [2018CVPR] Classification-Driven Dynamic Image Enhancement
• [2017CVPR] Forget Luminance Conversion and Do Something Better
• [2016CVPR] Two Illuminant Estimation and User Correction Preference
• Low-light Enhancement Repo [github]
• 基于深度学习的低光照图像增强方法总结(2017-2019) [Note]
• Learning to see, Antonio Torralba, 2016 [Slides]
• Attention-guided Low-light Image Enhancement, arXiv2019.8
• Low-light Image Enhancement Algorithm Based on Retinex and Generative Adversarial Network, arXiv2019.6
• LED2Net: Deep Illumination-aware Dehazing with Low-light and Detail Enhancement, arXiv2019.6
• EnlightenGAN: Deep Light Enhancement without Paired Supervision, arXiv2019.6 [Code]
• Kindling the Darkness: A Practical Low-light Image Enhancer, arXiv2019.5
• MSR-net: Low-light Image Enhancement Using Deep Convolutional Network, arXiv2017.11
• [2019TOG] Handheld Mobile Photography in Very Low Light
• [2019ICCV] Learning to See Moving Objects in the Dark
• [2019CVPR] Underexposed Photo Enhancement using Deep Illumination Estimation [Code]
• [2019CVPR] All-Weather Deep Outdoor Lighting Estimation
• [2019MMM] Progressive Retinex: Mutually Reinforced Illumination-Noise Perception Network for Low Light Image Enhancement
• [2018TIP] Learning a Deep Single Image Contrast Enhancer from Multi-Exposure Images
• [2018TMM] Naturalness preserved nonuniform illumination estimation for image enhancement based on retinex
• [2018PRL] LightenNet: A Convolutional Neural Network for weakly illuminated image enhancement
• [2018CVPR] Learning to See in the Dark
• [2018BMVC] MBLLEN: Low-light Image/Video Enhancement Using CNNs
• [2018BMVC] Deep Retinex Decomposition for Low-Light Enhancement
• [2018BMVC] Deep Retinex Decomposition for Low-Light Enhancement (Oral) [Proj] [Code]
• [2017TIP] LIME: Low-light image enhancement via illumination map estimation
• [2017PR] LLNet: A deep autoencoder approach to natural low-light image enhancement [Code] [Code2]
• [2017CVPR] Deep Outdoor Illumination Estimation
• [2016TIP] LIME: Low-light Image Enhancement via Illumination Map Estimation
• [2016ECCV] Deep Specialized Network for Illuminant Estimation

Reflection Removal

• [2019CVPR] Single Image Reflection Removal Beyond Linearity
• [2019CVPR] Reflection Removal Using A Dual-Pixel Sensor
• [2013ICCV] Exploiting Reflection Change for Automatic Reflection Removal

Denoising

• Deep Learning on Image Denoising: An overview, arXiv2020.1 [Proj]
• [2020NN] Attention-guided CNN for image denoising [Code]
• [2019CVPR] Toward Convolutional Blind Denoising of Real Photographs

Deblurring

• [SelfDeblur] Neural Blind Deconvolution Using Deep Priors, arXiv2019. 8 [Code]
• [2019ICCV] DeblurGAN-v2: Deblurring (Orders-of-Magnitude) Faster and Better [Code]
• [2018CVPR] DeblurGAN: Blind Motion Deblurring Using Conditional Adversarial Networks [Code]
• [2018CVPR] Dynamic Scene Deblurring Using Spatially Variant Recurrent Neural Networks

Deraining

• Single Image Deraining Rain Removal
• [2019CVPR] Single Image Deraining: A Comprehensive Benchmark Analysis
• [2018ECCV] Recurrent Squeeze-and-Excitation Context Aggregation Net for Single Image Deraining [Code]

Completion

• Image inpainting: A review, arXiv2019.9
• Consistent Generative Query Networks, arXiv2019.4 [Proj]
• [2019Scirobotics] Emergence of exploratory look-around behaviors through active observation completion [Proj]
• [2019ICCV] An Internal Learning Approach to Video Inpainting [Homepage] [Code] [Note]
• [2019ICCV] StructureFlow: Image Inpainting via Structure-aware Appearance Flow [Code]
• [2018Science] [GQN] Neural scene representation and rendering, DeepMind [Code] [Note]
• [2018CVPR] Learning to Look Around: Intelligently Exploring Unseen Environments for Unknown Tasks [Code]
• [2018CVPR] Deep Image Prior [github] [Note]
• [2018Proj] Painting outside the box: image outpainting with GANs, Mark Sabini, Stanford CS230 Project, arXiv2018.8 [Code] [PDF] [Model] [Note]

Image/Video Transfer

• Style Transfer Scholar: Dongdong Chen Dmitry Ulyanov

• [2018TOG] Progressive Color Transfer with Dense Semantic Correspondences ⭐️⭐️⭐️⭐️

• [2017CVPR] Improved Texture Networks: Maximizing Quality and Diversity in Feed-forward Stylization and Texture Synthesis

• [2016ICML] Texture Networks: Feed-forward Synthesis of Textures and Stylized Images [IN] [Code] [Slides]

• [2016CVPR] Image Style Transfer Using Convolutional Neural Networks, Gatys [Code]

• [2016ECCV] Perceptual Losses for Real-Time Style Transfer and Super-Resolution

• [2015] A neural algorithm of artistic style, Gatys, arXiv2015.9 [Code]

Blending/Fusion

• Deep Image Blending, arXiv201910 [Code]
• [2019MMM] GP-GAN: Towards Realistic High-Resolution Image Blending [Code] [Homepage]
• [2018ECCV] Learning to Blend Photos [Homepage]
• [2018SIGGA] Deep Blending for Free-Viewpoint Image-Based Rendering [Homepage]

Pedestrain/Crowd

PedestrainDetection

• Pedestrain Detection collection

• Deep Learning for Person Re-identification: A Survey and Outlook, arXiv2020.1 [Code]

• Pedestrain Attribute Recognition: A Survey, arXiv2019.1 [Proj]

• CrowdHuman: A Benchmark for Detecting Human in a Crowd, arXiv201804 [Proj] [Note]

• PedHunter: Occlusion Robust Pedestrian Detector in Crowded Scenes, arXiv2019.9

• [2020TMM/2019CVPRW] Bag of Tricks and A Strong Baseline for Deep Person Re-identification [Code]

• [2019ICCV] Mask-Guided Attention Network for Occluded Pedestrian Detection [Code]

• [2019CVPR] VRSTC: Occlusion-Free Video Person Re-Identification [occlusion]

• [2018CVPR] Repulsion Loss: Detecting Pedestrians in a Crowd, CVPR2018 [occlusion]

• [2016ECCV] Stacked Hourglass Networks for Human Pose Estimation

CrowdCounting

• awesome crowd counting

• Locate, Size and Count: Accurately Resolving People in Dense Crowds via Detection, arXiv2019.6 [Code]

• W-Net: Reinforced U-Net for Density Map Estimation, arXiv2019.3 [Unofficial Code]

• [2019TIP] HA-CCN: Hierarchical Attention-based Crowd Counting Network

• [2019ICCV] Bayesian Loss for Crowd Count Estimation with Point Supervision [Code]

• [2019ICCV] Crowd Counting with Deep Structured Scale Integration Network (oral) [github]

• [2019ICCV] Learning Spatial Awareness to Improve Crowd Counting (oral)

• [2019ICCV] Perspective-Guided Convolution Networks for Crowd Counting [Code] [Dataset]

• [2019ICCV] Learn to Scale: Generating Multipolar Normalized Density Maps for Crowd Counting

• [2019ICCV] Pushing the Frontiers of Unconstrained Crowd Counting: New Dataset and Benchmark Method

• [2019ICCV] Counting with Focus for Free [Code]

• [2019ICCVW] Crowd Counting on Images with Scale Variation and Isolated Clusters

• [2019CVPR] Learning from Synthetic Data for Crowd Counting in the Wild [Homepage] [Dataset]

• [2019MMM] Improving the Learning of Multi-column Convolutional Neural Network for Crowd Counting

• [2019ICME] Locality-constrained Spatial Transformer Network for Video Crowd Counting

• [2019SciAdvance] Number detectors spontaneously emerge ina deep neural network designed for visual object recognition [Note]

• [2019TII] Automated Steel Bar Counting and Center Localization with Convolutional Neural Networks [Code]

• [2018MICCAIW] Microscopy Cell Counting with Fully Convolutional Regression Networks [Code]

• [2010NIPS] Learning to count objects in images [Code]

GenerativeNet

GAN学习路线图：论文、应用、课程、书籍大总结 [Page]

深度学习中最常见GAN模型概览: GAN,DCGAN,CGAN,infoGAN,ACGAN,CycleGAN,StackGAN ...

Blog: One Day One GAN

Training Tricks

• How to Train a GAN? Tips and tricks to make GANs work [Page]

  Start from NIPS2016, 17 GAN tricks, by Soumith Chintala, Emily Denton, Martin Arjovsky, Michael Mathieu. How to Train a GAN, NeurIPS2016

• Top highlight Advances in Generative Adversarial Networks (GANs): A summary of the latest advances in Generative Adversarial Networks [Page] [Note]

• Keep Calm and train a GAN. Pitfalls and Tips on training Generative Adversarial Networks [Page]

• Image Augmentations for GAN Training. arXiv202006

Papers

• [Blogg] A Beginner's Guide to Generative Adversarial Networks (GANs), 2019

• Generative Adversarial Networks: A Survey and Taxonomy, arXiv2020.2 [GANReview]

• A Review on Generative Adversarial Networks: Algorithms, Theory, and Applications, arXiv202001

• [2019ACMCS] How Generative Adversarial Networks and Their Variants Work: An Overview

• StarGAN v2: Diverse Image Synthesis for Multiple Domains. arXiv201912 [Code]

• This dataset does not exist: training models from generated images, arXiv2019.11

• Landmark Assisted CycleGAN for Cartoon Face Generation. arXiv201907

• Maximum Entropy Generators for Energy-Based Models, arXiv2019.5 [Code]

• [2019NIPS] Few-shot Video-to-Video Synthesis [Code]

• [2019NIPS] [vid2vid] Video-to-Video Synthesis [Code]

• [2019CVPR] Semantic Image Synthesis with Spatially-Adaptive Normalization [Proj] [Code]

• [2019CVPR] [seg2vid] Video Generation from Single Semantic Label Map [Code]

• [2019BMVC] The Art of Food: Meal Image Synthesis from Ingredients

• [2018ICLR] Spectral Normalization for Generative Adversarial Networks [Code] [Supp1] [Supp2]

• [2018CVPR] [pix2pixHD] High-Resolution Image Synthesis and Semantic Manipulation with Conditional GANs

• [2018CVPR] StarGAN: Unified Generative Adversarial Networks for Multi-Domain Image-to-Image Translation (oral) [Code]

• [2018ECCV] [FE-GAN] Fashion Editing with Multi-scale Attention Normalization [Notes]

• [2018ECCV] Image Inpainting for Irregular Holes Using Partial Convolutions [Code] [Code2] [used for DeepNude]

• [2017ICCV] [CycleGAN] Unpaired Image-to-Image Translation using Cycle-Consistent Adversarial Networks [Proj]

• [2017CVPR] [Pix2Pix] Image-to-Image Translation with Conditional Adversarial Networks [Demo]

• [2016ICLR] [DCGAN] Unsupervised representation learning with deep convolutional generative adversarial networks

• [2016ICML] A Theory of Generative ConvNet [S-C Zhu] [Proj/Code]

• [2014NIPS] Generative Adversarial Nets

Video Understanding

• [YOWO] You Only Watch Once: A Unified CNN Architecture for Real-Time Spatiotemporal Action Localization arXiv201911 [Code]

• [2019CVPR] Learning Video Representations from Correspondence Proposals

  现有的视频深度学习架构通常依赖于三维卷积、自相关、非局部模块等运算，这些运算难以捕捉视频中帧间的长程运动/相关性，该文提出的CPNet学习视频中图片之间的长程对应关系，来解决现有方法在处理视频长程运动中的局限性.

Video Object Detection

• Object Detection in Video with Spatial-temporal Context Aggregation, arXiv2019.7
• Looking Fast and Slow: Memory-Guided Mobile Video Object Detection, arXiv2019.3 [TF] [PyTorch]
• [2019ICCV] [MGAN] Motion Guided Attention for Video Salient Object Detection
• [2019CVPR] Shifting More Attention to Video Salient Objection Detection [Code]
• [2019CVPR] Activity Driven Weakly Supervised Object Detection [Code]
• [2019SysML] AdaScale: Towards Real-time Video Object Detection Using Adaptive Scaling
• [2019KDDW] Understanding Video Content: Efficient Hero Detection and Recognition for the Game "Honor of Kings" [Notes]](https://flashgene.com/archives/28803.html)
• [2018CVPR] Mobile Video Object Detection With Temporally-Aware Feature Maps

Video Object segmentation

• [2016CVPR] A Benchmark Dataset and Evaluation Methodology for Video Object Segmentation

• [2019ICCV] RANet: Ranking Attention Network for Fast Video Object Segmentation

• [2019CVPR] See More, Know More: Unsupervised Video Object Segmentation with Co-Attention Siamese Networks [Code]

• [2019CVPR] Improving Semantic Segmentation via Video Propagation and Label Relaxation [Code]

DeepCNN

• Deep Learning Papers Reading Roadmap

Optimization

Summary of SGD，AdaGrad，Adadelta，Adam，Adamax，Nadam

Why Momentum Really Works, 2017

• Optimization for deep learning: theory and algorithms. arXiv201912 [[OptimizationCourse]](Optimization Theory for Deep Learning)

• Why Adam Beats SGD for Attention Models. arXiv201912

• Momentum Contrast for Unsupervised Visual Representation Learning, Kaiming He arXiv2019.11

• Dynamic Mini-batch SGD for Elastic Distributed Training: Learning in the Limbo of Resources, Amazon, arXiv2019.5

• Accurate, Large Minibatch SGD: Training ImageNet in 1 Hour, arXiv2018.4 [Notes]

• [2019NIPS] Uniform convergence may be unable to explain generalization in deep learning

• [2019NIPS] Understanding the Role of Momentum in Stochastic Gradient Methods

• [2019NIPS] Lookahead optimizer: k steps forward, 1 step back [Code] [Pytorch] [TF]

• [2019ICLR] [AdaBound] Adaptive gradient methods with dynamic bound of learning rate [Pytorch] [TF-example]

  AdaBound combines SGD and Adam to make it fast as Adam at training start and convergence like SGD later. Usage: require Python 3.6+, and pip install: pip install adabound, and then: optimizer = adabound.AdaBound(model.parameters(), lr=1e-3, final_lr=0.1). Version of TensorFlow is coming.

• [2019CVPRW] The Indirect Convolution Algorithm

• [2019ISCAW] Accelerated CNN Training Through Gradient Approximation

Training

Fast training for neural networks, You Yang, Jiangmen Talk [Video]

Training Tricks in Object Detection

• Student Specialization in Deep ReLU Networks With Finite Width and Input Dimension, arXiv2019.11
• Accelerating CNN Training by Sparsifying Activation Gradients, arXiv2019.8
• Luck Matters: Understanding Training Dynamics of Deep ReLU Networks, arXiv2019.6 [Code]
• Bag of Freebies for Training Object Detection Neural Networks, Amazon, arXiv2019.4 [[Code]](https://github.com/dmlc/gluon-cv\)
• Deep Double Descent: Where Bigger Models and More Data Hurt, ICLR2020Review
• [2019ICCV] Rethinking ImageNet Pre-training, FAIR [Notes]
• [2019CVPR] Bag of Tricks for Image Classification with Convolutional Neural Networks, Amazon [Code] [Note]
• [2019CVPR] Accelerating Convolutional Neural Networks via Activation Map Compression
• [2019CVPR] RePr: Improved Training of Convolutional Filters [Note]
• [2019BMVC] Dynamic Neural Network Channel Execution for Efficient Training
• [2018ICPP] Imagenet training in minutes

Activation

• [Blog] 深度学习中的激活函数

  Dead Relu [Notes]

• [2019CVPR] Why ReLU networks yield high-confidence predictions far away from the training data and how to mitigate the problem (oral) [Code]

• [2018] [GELU] Gaussian Error Linear Units (GELUs). arXiv201811 [Note]

  # GELU in GPT-2: def gelu(x): return 0.5*x*(1+tf.tanh(np.sqrt(2/np.pi)*(x+0.044715*tf.pow(x, 3))))
  

• [2016ICML] [CReLU] Understanding and improving convolutional neural networks via concatenated rectified linear units

• [2015ICCV] [PReLU-Net/msra Initilization] Delving Deep into Rectifiers: Surpassing Human-Level Performance on ImageNet Classification

Normalization

• Normalization Scholar: Ping Luo

  [Blog] Introduction to Normalization [Page] [Note]

  [Blog] Introduction to BN/LN/IN/GN [Page] [Page2]

  [Talk] Devils in BatchNorm, Jiangmen Talk, 2019 [Page]

  [Blog] An Overview of Normalization Methods in Deep Learning, 2018.11 [Page]

• Attentive Normalization. [Tianfu Wu] arXiv2019.11 [Code]

• Network Deconvolution. [a alternative to Batch Normalization]. arXiv2019.9 [Proj]

• Weight Standardization. arXiv2019.3 [Code]

• [IN] Instance Normalization: The Missing Ingredient for Fast Stylization. arXiv2017.11 [Code]

• [LN] Layer Normalization. [Hinton] arXiv2016.7 [Note]

• [2019NIPS] Understanding and Improving Layer Normalization

• [2019NIPS] Positional Normalization [Code] [Supp]

• [2018NIPS] How Does Batch Normalization Help Optimization? [arXiv19v] [Ref]

• [2018NIPS] [BIN] Batch-Instance Normalization for Adaptively Style-Invariant Neural Networks [Code]

• [2018ECCV] [GN] Group normalization

• [2017NIPS] Batch Renormalization: Towards Reducing Minibatch Dependence in Batch-Normalized Models

• [2016NIPS] [WN] Weight normalization: A simple reparameterization to accelerate training of deep neural networks

• [2015ICML] [BN] Batch Normalization: Accelerating Deep Network Training by Reducing Internal Covariate Shift

Dropout

• Dropout [Note1] [Note2]

• [2014JMLR] Dropout: a simple way to prevent neural networks from overfitting

• [2012NIPS] ImageNet Classification with Deep Convolutional Neural Networks

Augmentation

• [Blog] Research Guide: Data Augmentation for Deep Learning. 201910

  [Blog] Data Augmentation: How to use Deep Learning when you have Limited Data. 201805 [Page]

• [2019JBD] A survey on Image Data Augmentation for Deep Learning. [PDF] [Notes]

• Data Augmentation Revisited: Rethinking the Distribution Gap between Clean and Augmented Data. arXiv2019.11

• FMix: Enhancing Mixed Sample Data Augmentation arXiv202006 [Code]

• GridMask Data Augmentation. arXiv202001 [Code] [Note]

• Let’s Get Dirty: GAN Based Data Augmentation for Soiling and Adverse Weather Classification in Autonomous Driving. arXiv2019.12

• PanDA: Panoptic Data Augmentation, arXiv2019.11

• Faster AutoAugment: Learning augmentation strategies using backpropagation. arXiv201911

• Automatic Data Augmentation by Learning the Deterministic Policy. arXiv201910

• Greedy AutoAugment, arXiv2019.8

• Safe Augmentation: Learning Task-Specific Transformations from Data, arXiv2019.7 [Code]

• Learning Data Augmentation Strategies for Object Detection. arXiv201906 [Code]

• [2020ICLR] AugMix: A Simple Data Processing Method to Improve Robustness and Uncertainty [Code]

• [2019NIPS] Implicit Semantic Data Augmentation for Deep Networks

• [2019NIPS] Fast AutoAugment

• [2019ICML] Population Based Augmentation: Efficient Learning of Augmentation Policy Schedules [Code] [Examples]

• [2019ICCV] CutMix: Regularization Strategy to Train Strong Classifiers with Localizable Features [Code]

• [2019ICCVW] Occlusions for Effective Data Augmentation in Image Classification

• [2019ICCVW] Style Augmentation: Data Augmentation via Style Randomization

• [2019CVPR] AutoAugment: Learning Augmentation Policies from Data [Code]

• [2018ICLR] Mixup: Beyond empirical risk minimization

• [2018ACML] RICAP: Random Image Cropping and Patching Data Augmentation for Deep CNNs [Code]

• [2018ICANN] Further advantages of data augmentation on convolutional neural networks (best paper)

Model

[Blog] 从Softmax到AMSoftmax

[Blog] Convolutional Neural Networks Structure

[Blog] A Survey of the Recent Architectures of Deep Convolutional Neural Networks, 2019

[Blog] CNN下/上采样详析

• A closer look at network resolution for efficient network design. arXiv201909 [Code]
• [2019NIPS] Is Deeper Better only when Shallow is Good? [Code]
• [2015Nature] Deep Learning Review
• [2014BMVC] Return of the Devil in the Details: Delving Deep into Convolutional Nets

Module

• Pooling:

  ViP: Virtual Pooling for Accelerating CNN-based Image Classification and Object Detection, arXiv201906

  Learning Spatial Pyramid Attentive Pooling in Image Synthesis and Image-to-Image Translation, arXiv201901

  [2020AAAI] Revisiting Bilinear Pooling: A coding Perspective [Note]

  [2019ICCV] LIP: Local Importance-based Pooling [Code] [Notes]

  [2018ECCV] Grassmann Pooling as Compact Homogeneous Bilinear Pooling for Fine-Grained Visual Classification

  [2017CVPR] Low-rank bilinear pooling for fine-grained classification

  [2016EMNLP] Multimodal compact bilinear pooling for visual question answering and visual grounding

  [2016CVPR] Compact bilinear pooling

  [2015ICCV] [bilinear pooling] Bilinear CNN Models for Fine-grained Visual Recognition

  [2012ECCV] Semantic segmentation with second-order pooling

• Dynamic Convolutions: Exploiting Spatial Sparsity for Faster Inference arXiv201912

• Rethinking Softmax with Cross-Entropy: Neural Network Classifier as Mutual Information Estimator, arXiv201911

• Rethinking the Number of Channels for the Convolutional Neural Network, arXiv201909

• AutoGrow: Automatic Layer Growing in Deep Convolutional Networks, arXiv201909 [Code]

• Mapped Convolutions. [For 2D/3D/Spherical]. arXiv201906 [Code]

• Spatial Group-wise Enhance: Enhancing Semantic Feature Learning in Convolutional Networks. arXiv201905 [Code] [Note]

• [2019ICCV] ACNet: Strengthening the Kernel Skeletons for Powerful CNN via Asymmetric Convolution Blocks [Code]

• [2019CVPRW] Convolutions on Spherical Images

• [2017ICML] Warped Convolutions: Efficient Invariance to Spatial Transformations

• Attention module

• ECA-Net: Efficient Channel Attention for Deep Convolutional Neural Networks, arXiv201910 [Code] [Chinese]

• [2020ICLR] On the Relationship between Self-Attention and Convolutional Layers [Proj] [Code] [Intro]

• [2019TIP] Region Attention Networks for Pose and Occlusion Robust Facial Expression Recognition [Code]

• [2017CVPR] SCA-CNN: Spatial and Channel-wise Attention in Convolutional Networks for Image Captioning

Backbone

• Comb Convolution for Efficient Convolutional Architecture. arXiv201911
• [2019ICML] EfficientNet: Rethinking Model Scaling for Convolutional Neural Networks [Code]
• [2019ICCVW] GCNet: Non-local Networks Meet Squeeze-Excitation Networks and Beyond [Code]
• [2018CSVT] [RoR] Residual Networks of Residual Networks: Multilevel Residual Networks
• [2018CVPR] [SENet] Squeeze-and-excitation networks
• [2017ICLR] FractalNet: Ultra-Deep Neural Networks without Residuals
• [2017CVPR] [PyramidNet] Deep Pyramidal Residual Networks
• [2017CVPR] [DenseNet] Densely Connected Convolutional Networks
• [2017CVPR] [ResNeXt] Aggregated Residual Transformations for Deep Neural Networks
• [2017CVPR] Xception: Deep Learning with Depthwise Separable Convolutions
• [2017CVPR] PolyNet: A Pursuit of Structural Diversity in Very Deep Networks [Slides]
• [2017AAAI] Inception-v4, Inception-ResNet and the Impact of Residual Connections on Learning
• [2016CVPR] [ResNet] Deep Residual Learning for Image Recognition [Note1] [Note2]
• [2016CVPR] [Inception-v3] Rethinking the Inception Architecture for Computer Vision
• [2016ECCV] Good Practices for Deep Feature Fusion
• [2016ECCV] Deep Networks with Stochastic Depth
• [2016ECCV] [Identity ResNet] Identity Mappings in Deep Residual Networks [Over 1000 Layers ]
• [2016ICLRW] ResNet in ResNet: Generalizing Residual Architectures
• [2015NIPS] [STN] Spatial Transformer Networks
• [2015ICML] [BN-Inception /Inception-v2] Batch Normalization: Accelerating Deep Network Training by Reducing Internal Covariate Shift
• [2015CVPR] [GoogLeNet/Inception-v1] Going Deeper with Convolutions
• [2015ICLR] [VGGNet] Very Deep Convolutional Networks for Large-Scale Image Recognition
• [2014ICLR] [NIN] Network in Network
• [2014ECCV] [ZFNet] Visualizing and Understanding Convolutional Networks
• [2014MMM] [CaffeNet] Caffe: Convolutional Architecture for Fast Feature Embedding
• [2012NIPS] [AlexNet] Imagenet classification with deep convolutional neural networks
• [1998ProcIEEE] [LeNet] Gradient-Based Learning Applied to Document Recognition [LeNet Notes]

Light-weightCNN

• [Blog] Introduction of light-weight CNN

  [Blog] Lightweight convolutional neural network: SqueezeNet、MobileNet、ShuffleNet、Xception

• SeesawNet: Convolution Neural Network With Uneven Group Convolution. arXiv201912 [Code]

• HGC: Hierarchical Group Convolution for Highly Efficient Neural Network, arXiv201906

• [2020CVPR] GhostNet: More Features from Cheap Operations [Code]

• [2019CVPR] ESPNetv2: A Light-weight, Power Efficient, and General Purpose Convolutional Neural Network [Code]

  [2018ECCV] ESPNet: Efficient Spatial Pyramid of Dilated Convolutions for Semantic Segmentation

• [2019CVPRW] Depth-wise Decomposition for Accelerating Separable Convolutions in Efficient Convolutional Neural Networks

• [2019BMVC] MixNet: Mixed Depthwise Convolutional Kernels [Code] [Notes]

• [2018NIPS] ChannelNets: Compact and Efficient Convolutional Neural Networks via Channel-Wise Convolutions [Code]

• [2018NIPS] Learning Versatile Filters for Efficient Convolutional Neural Networks [Code]

• [2018BMVC] IGCV3: Interleaved Low-Rank Group Convolutions for Efficient Deep Neural Networks [Code] [Pytorch]

  [2018CVPR] IGCV2: Interleaved Structured Sparse Convolutional Neural Networks

  [2017ICCV] [IGVC1] Interleaved Group Convolutions for Deep Neural Networks

• MobileNet Series:

  [Blog] Introduction for MobileNet and Its Variants

  [2019ICCV] Searching for MobileNetV3. [Note]

  [2018CVPR] MobileNetV2: Inverted Residuals and Linear Bottlenecks. [Note]

  [2017] MobileNets: Efficient Convolutional Neural Networks for Mobile Vision Applications. arXiv201704

• ShuffleNet Series [Note]

  [Code] ShuffleNet Series by Megvii: ShuffleNetV1, V2/V2+/V2.Large/V2.ExLarge, OneShot, DetNAS

  [2018ECCV] ShuffleNet V2: Practical Guidelines for Efficient CNN Architecture Design

  [2018CVPR] ShuffleNet: An Extremely Efficient Convolutional Neural Network for Mobile Devices

Interpretation

[Blog] 深度神经网络可解释性方法汇总(附TF代码实现)

• Analysis of Explainers of Black Box Deep Neural Networks for Computer Vision: A Survey. arXiv2019.11
• [2019NIPS] Wide Neural Networks of Any Depth Evolve as Linear Models Under Gradient Descent [Code] [Note]
• [2019NIPS] Weight Agnostic Neural Networks (spotlight). [Proj] [Note]
• [2018AAAI] Interpreting CNN Knowledge via An Explanatory Graph
• [2018Acces] Peeking Inside the Black-Box: A Survey on Explainable Artificial Intelligence (XAI)

Others

• RetinaFace: Single-stage Dense Face Localisation in the Wild. arXiv201905 [Code-MXNet] [Code-TF]

• [2019CVPR] Group Sampling for Scale Invariant Face Detection [Note]

• [2019ICCV] Learning to Paint with Model-based Deep Reinforcement Learning [Code] [Note]

• [2019ICCV] Fashion++: Minimal Edits for Outfit Improvement (FAIR) [Proj] [Code]

• [2019ICCV] SpatialSense: An Adversarially Crowdsourced Benchmark for Spatial Relation Recognition [Code&Dataset]

• [2018BMVC] Learning Geo-Temporal Image Features [Proj]

AI+Music

• [2018ISMIR] MIDI-VAE: Modeling Dynamics and Instrumentation of Music with Applications to Style Transfer [Code]
• Music continue: MuseNet Bach-AI-Music-Google Generating Piano Music with Transformer by Google
• On the Measure of Intelligence. arXiv201911 [Intro]

Unsupervised Learning

• A Simple Framework for Contrastive Learning of Visual Representations. arXive202002

Pose

• Pose Trainer: Correcting Exercise Posture using Pose Estimation, arXiv github

AI+Application

• MetNet: A Neural Weather Model for Precipitation Forecasting, arXiv202003 [Blog] Intro