hadryan / recent-stars-2022

🔥🔥🔥SLAM, VIsual localization, keypoint detection, Image matching, Pose/Object tracking, Depth/Disparity/Flow Estimation, 3D-graphic, etc. related papers and code

Recent Stars 2022

✔ This repo collects some links with papers which I recently starred related on SLAM, Pose/Object tracking, Depth/Disparity/Flow Estimation, 3D-graphic, etc.

近期写了一个自动获取Arxiv上有关SLAM/特征提取匹配/视觉定位等领域论文的小工具cv-arxiv-daily，每日更新，欢迎大家关注。 Recommend: A useful tool to automatically update CV papers daily using github actions (Update Every day)

SLAM related

最近主要关注视觉定位+SFM算法（Last Update: 2021.12.31）

• [Light Flow] Learning Optical Flow from a Few Matches, CVPR 2021, [PDF], 光流估计

• [SFM] Beyond Controlled Environments: 3D Camera Re-Localization in Changing Indoor Scenes, ECCV 2020, [PDF], 已知2D-3D关联和相对位姿，用于多相机位姿估计

• [SLAM] Tangent Space Backpropagation for 3D Transformation Groups, CVPR 2021, [PDF], 使用Torch实现李群反向传播

• [Localization] Visual-Based-Localization-Papers, 视觉定位相关论文集

• [Localization] kapture-localization: toolbox, 基于kapture的视觉定位流程

• [Localization] SeqNet: Learning Descriptors for Sequence-based Hierarchical Place Recognition, ICRA 2021, [PDF],[Video], 一种基于图像序列的场景识别算法

• [Event Camera] DSEC: A Stereo Event Camera Dataset for Driving Scenarios, CVPRW 2021, [PDF],[Homepage], 自动驾驶环境下，使用事件相机+双目全局曝光+LIDAR+RTK GPS采集的数据集

• [SFM] openMVS_comments, OpenMVS 注释版

• [3D Reconstruction] J3DReconstruction, Windows下基于openMVG+openMVS的三维重建解决方案以及基于Qt的可视化桌面平台

• [PointCloud] 3D-PointCloud, 点云相关论文以及数据集

• [SLAM] VIDO-SLAM, 单目相机紧耦合动态物体环境下SLAM

• [SLAM] Interactive Visualisation of Gaussian processes, [Homepage], 高斯过程动画演示

• [SFM] ROBA: Rotation-Only Bundle Adjustment, CVPR 2021, [PDF],[Video], 高效优化旋转量，用于全局SFM，提高SFM精度

• [Localization] Fusion-DHL: WiFi, IMU, and Floorplan Fusion for Dense History of Locations in Indoor Environments, ICRA 2021, [PDF],[Video], 多模态（WiFi,IMU,Floorplan）传感器数据融合室内定位

• [SLAM] VINS-GPS-Wheel, 基于VINS-Mono开发的SLAM算法，轮式紧耦合，GPS松耦合

• [SLAM] OpenREALM,[Video], 基于OpenVSLAM开发的一套开源SLAM框架，可用于实时无人机建图定位

• [Localization] InLoc_demo, 用于验证室内数据集InLoc视觉定位效果的演示脚本

• [L-SLAM] T-LOAM: Truncated Least Squares Lidar-only Odometry and Mapping in Real-Time, TGRS 2021, [PDF],[Video], 使用截断最小二乘+Open3D点云库实现SLAM

• [SLAM] ADEKF, 无需定义雅可比，ceres实现自动微分卡尔曼滤波

• [L-SLAM] Locus: LiDAR-based Place Recognition using Spatiotemporal Higher-Order Pooling, ICRA 2021, [PDF],[Homepage], 3D点云全局描述子实现场景识别

• [Matching] COTR: Correspondence Transformer for Matching Across Images, CVPR 2021, [PDF],[Homepage], 基于Transformer的图像匹配，非特征点也可匹配

• [Matching] NRE: Neural Reprojection Error: Merging Feature Learning and Camera Pose Estimation, CVPR 2021, [PDF],[Homepage], 改进传统的重投影误差，使用深度学习重新设计重投影误差的形式，实现特征匹配和相机位姿估计

• [Matching]Learning Accurate Dense Correspondences and When to Trust Them, CVPR 2021 (Oral), [PDF],稠密特征匹配

• [SFM]How privacy preserving are Line Clouds? Recovering Scene Details from 3D Lines, CVPR 2021, [PDF],[Video],隐私保护SFM将模型中的点转换成了线，视觉（人）根本看不出场景的原本的样子，本文反向将这些线转换成了点 ，进而恢复场景的结构。

• [L-SLAM]R2LIVE: A Robust, Real-time, LiDAR-Inertial-Visual tightly-coupled state Estimator and mapping, arxiv 2020, [PDF],[Video],激光+IMU+视觉紧耦合SLAM

• [SFM] InvSFM: Revealing Scenes by Inverting Structure from Motion Reconstructions, CVPR 2019 (Oral), [PDF],[Homepage], 反向SFM，从点云恢复场景

• [Feature] LETR: Line Segment Detection Using Transformers without Edges, CVPR2021 (Oral), [PDF], 利用Transformer实现端到端线段提取

• [Feature] FFD: Fast Feature Detector, TIP 2020, [PDF], [Blog]，传统方式实现快速特征提取器，该方法的复杂度小于目前流行SIFT约5%

• [Feature] Learning and aggregating deep local descriptors for instance-level recognition, ECCV 2020, [PDF], 深度局部描述子实现instance-level的识别

• [Mapping] Multi-View Optimization of Local Feature Geometry, ECCV 2020, [PDF],[Homepage]，多视角优化2D点进而提高SFM建图的精度，如降低平均重投影误差

• [Localization&Mapping] Cross-Descriptor Visual Localization and Mapping, arxiv 2020, [PDF], 跨描述子视觉建图与定位：能够使建图特征进行“更新”以及实现了跨特征类型的匹配（如使用SIFT建立的场景模型，可以用HardNet进行定位）

• [SLAM] CamVox: A Low-cost and Accurate Lidar-assisted Visual SLAM System, arxiv 2020, [PDF]，[Blog]，相机和激光雷达融合（SLAM）

• [Localization] Stereo Localization in LiDAR Maps, 跨模态定位：在LiDAR地图中使用视觉图像定位

• [Localization] Augmenting Visual Place Recognition with Structural Cues, Robotics and Automation Letters (RA-L) 2020, [PDF]，[Homepage]，结合外观以及3D结构线索增强视觉定位，目前效果远超NetVLad

• [Localization]CMRNet: Camera to LiDAR-Map Registration, PDF: [CMRNet, ITSC 2019], [CMRNet++, ICRA 2020], [Homepage]，在LIDAR地图中用RGB定位，以初始位姿开始，迭代出定位位姿

• [Localization]AtLoc: Attention Guided Camera Localization, AAAI 2020, [PDF]，注意力机制视觉定位

• [Localization]Hierarchical-Localization, PDF, [From Coarse to Fine: Robust Hierarchical Localization at Large Scale,CVPR 2019], [SuperGlue: Learning Feature Matching with Graph Neural Networks, CVPR 2020], 目前视觉定位挑战赛visuallocalization.net/benchmark TOP 1的算法（使用了Hierarchical Localization - SuperPoint + SuperGlue）。

• [Localization]Kapture: Robust Image Retrieval-based Visual Localization using Kapture, arXiv 2020, [PDF], 基于3D模型的视觉定位，局部特征支持R2D2,D2-Net,全局特征为AP-GeM,另外提出了一种灵活的数据组织格式Kapture，能够轻易地支持导入/出数据到现有的SfM软件

• [Localization]CamOdoCal: Automatic Intrinsic and Extrinsic Calibration of a Rig with Multiple Generic Cameras and Odometry, RSJ International Conference on Intelligent Robots & Systems 2013, [PDF]

• [Localization]Night-to-Day Image Translation for Retrieval-based Localization, arXiv 2018, [PDF]， 黑夜转白天准确视觉定位

• [Localization]DSAC: DSAC – Differentiable RANSAC for Camera Localization, CVPR 2017, [PDF], [Homepage]

• [Localization]ESAC: Expert Sample Consensus Applied to Camera Re-Localization, ICCV 2019, [PDF], [Homepage]

• [Localization]DIFL-FCL：Retrieval-based Localization Based on Domain-invariant Feature Learning under Changing Environments, IROS 2019, [PDF]

• [Localization]Visual Localization Under Appearance Change: A Filtering Approach, DICTA 2019, [PDF]

• [Localization]A Generative Map for Image-based Camera Localization, 2019, [PDF]，视觉定位

• [Localization]DISAM: Domain-invariant Similarity Activation Map Metric Learning for Retrieval-based Long-term Visual Localization, IROS 2019, [PDF]，基于图像召回的视觉定位

• [Localization]RGB2LIDAR: Towards Solving Large-Scale Cross-Modal Visual Localization, ACM MM 2020, [PDF]，在LIDAR地图中用RGB定位

• [Localization]Multi-Process Fusion: Visual Place Recognition Using Multiple Image Processing Methods, IEEE RAL 2019, [PDF]

• [Localization]Learning Two-View Correspondences and Geometry Using Order-Aware Network, ICCV 2019, [PDF]

• [L-SLAM]LeGO-LOAM: Lightweight and Ground-Optimized Lidar Odometry and Mapping on Variable Terrain, IROS 2018, [PDF]

• [SfM]Multi-View Optimization of Local Feature Geometry, ECCV 2020, [PDF], [Homepage], [Video]

• [VIO]Robust and Efficient Visual-Inertial Odometry with Multi-plane Priors, PRCV 2019, [PDF]， 多平面先验VI里程计

• [Relocalization]Online Visual Place Recognition via Saliency Re-identification, IROS 2020, [PDF], [Homepage]

• [SLAM]DXSLAM: A Robust and Efficient Visual SLAM System with Deep Features, arXiv 2020, [PDF]

• [Feature]Learning Feature Descriptors using Camera Pose Supervision, ECCV 2020, [PDF], [Homepage]

• [Feature]Efficient adaptive non-maximal suppression algorithms for homogeneous spatial keypoint distribution,Pattern Recognition Letters 2019，特征点平均分布

• [VIO]ALVIO: Adaptive Line and Point Feature-based Visual Inertial Odometry for Robust Localization in Indoor Environments, 点+线特征

• [SLAM]ORB-SLAM3: An Accurate Open-Source Library for Visual, Visual-Inertial and Multi-Map SLAM, [PDF]

• [SLAM]LIO-SAM, 激光雷达IMU紧耦合SLAM

• [Tool]Robotics Toolbox for Python, a Python implementation of the Robotics Toolbox for MATLAB®

• [Matching]LISRD,ECCV 2020, [PDF]，在线局部不变特征匹配！重要！

• [Matching]AdaLAM,特征匹配快速滤除外点

• [Calib]fisheye_pinhole_calib_demo, 包括鱼眼模型、针孔模型的相机标定，封装了自动编译、库的打包以及外部库的调用测试

• [Calib]SensorCalibration, IMU雷达标定

• [VO]Low-Drift Visual Odometry in Structured Environments by Decoupling Rotational and Translational Motion,ICRA 2018, [PDF], 结构化环境中将旋转量与平移量进行分离优化

• [VIO]VIO-SLAM, 从零开始手写VIO课后作业

• [Matching]TFMatch: Learning-based image matching in TensorFlow,TensorFlow 实现的 GeoDesc,ASLFeat以及ContextDesc

• [Tutorial]SLAM-BOOK, 一本关于SLAM的书稿，清楚的介绍SLAM系统中的使用的几何方法和深度学习方法，持续更新中

• [Loop Closing]OverlapNet - Loop Closing for 3D LiDAR-based SLAM, RSS 2020, [PDF], 3D激光雷达SLAM闭环

• [SLAM]VDO_SLAM, RGB-D相机数据作为输入，实现追踪动态物体SLAM的功能, [PDF]

• [SLAM]orbslam-map-saving-extension，在ORB-SLAM的基础上增加保存+加载地图功能

• [Tutorial]Modern Robotics: Mechanics, Planning, and Control Code Library, 现代机器人学, [Homepage]

• [Matching]image-matching-benchmark-baselines, 图像特征匹配挑战赛主页

• [Matching]GraphLineMatching

• [Matching]Locality Preserving Matching, IJCAI 2017, [PDF]

• [IMU]IMUOrientationEstimator

• [Feature]BEBLID: Boosted Efficient Binary Local Image Descriptor

• [Relocalization]KFNet: Learning Temporal Camera Relocalization using Kalman Filtering,CVPR 2020,[PDF]

• [Matching]image-matching-benchmark

• [Matching]GMS: Grid-based Motion Statistics for Fast, Ultra-robust Feature Correspondence,CVPR 17 & IJCV 19,[PDF],[Project page]

• [Reloc]GN-Net-Benchmark, CVPR 2020,GN-Net: The Gauss-Newton Loss for Multi-Weather Relocalization, [PDF],[Project page]

• [Matching]SuperGluePretrainedNetwork, CVPR 2020, [PDF], 划重点！2020年sota超大视角2D特征匹配，Blog

• [Feature]D3Feat, CVPR 2020, [PDF]

• [Feature]ASLFeat, CVPR 2020, ASLFeat: Learning Local Features of Accurate Shape and Localization, [PDF]

• [Feature]GMS-Feature-Matcher, CVPR 2018, GMS: Grid-based Motion Statistics for Fast, Ultra-robust Feature Correspondence, [PDF],[Project page]

• [Feature]D3Feat, CVPR 2020, [PDF]

• [Feature]3DFeatNet, ECCV 2018, [PDF]

• [Tutorial]AutonomousDrivingCookbook，Scenarios, tutorials and demos for Autonomous Driving

• [Tutorial]SLAMPaperReading，泡泡机器人北京线下SLAM论文分享资料

• [Tutorial]VIO_Tutotial_Course

• [Tutorial]VO-SLAM-Review

• [Tutorial]VINS-Mono-code-annotation,VINS-Mono代码注释以及公式推导

• [Tutorial]VINS-Mono-Learning,VINS-Mono代码注释

• [Tutorial]VINS-Course,VINS-Mono code without Ceres or ROS

• [Tutorial]VIO-Doc,主流VIO论文推导及代码解析

• [VO]CNN-DSO, Direct Sparse Odometry with CNN Depth Prediction

• [VO]fisheye-ORB-SLAM, A real-time robust monocular visual SLAM system based on ORB-SLAM for fisheye cameras, without rectifying or cropping the input images

• [VO]ORB_Line_SLAM, Real-Time SLAM with BoPLW Pairs for Stereo Cameras, with Loop Detection and Relocalization Capabilities

• [VO]DeepVO-pytorch, ICRA 2017 DeepVO: Towards end-to-end visual odometry with deep Recurrent Convolutional Neural Networks

• [Calib]CamOdomCalibraTool, The tool to calibrate extrinsic param between camera and wheel.

• [Calib]lidar_camera_calibration,another version

• [Calib]OdomLaserCalibraTool，相机与2D雷达标定

• [Calib]extrinsic_lidar_camera_calibration, LiDARTag: A Real-Time Fiducial Tag using Point Clouds, arXiv 2019, [PDF]

• [Calib]velo2cam_calibration, Automatic Calibration algorithm for Lidar-Stereo camera, [Project page]

• [Dataset]IRS: A Large Synthetic Indoor Robotics Stereo Dataset for Disparity and Surface Normal Estimation

• [Tools]averaging-quaternions,四元数平均

---

分割线，以下是2019年的星标项目，上面是2020年新星标的。

• R2D2: Reliable and Repeatable Detector and Descriptor,NeurIPS 2019,[PDF],[Project page]，深度学习特征点+描述子

• Semantic_SLAM,语义SLAM：ROS + ORB SLAM + PSPNet101

• PlaceRecognition-LoopDetection, Light-weight place recognition and loop detection using road markings

• DOOR-SLAM: Distributed, online, and outlier resilient SLAM for robotic teams,[PDF],[Project page]，多机器人协作SLAM，增强了场景的适用性

• awesome-local-global-descriptor, 超详细深度学习特征点描述子集合，需要重点关注一下这个repo

• GIFT: Learning Transformation-Invariant Dense Visual Descriptors via Group CNNs, NeurIPS 2019，[PDF], [Project page]，浙大CAD+商汤联合实验室出品，利用Group CNN来改进superpoint描述子（仅描述，特征点提取可任意选择），可以大幅度增强视角变化时的特征点复检率与匹配点数

• Key.Net: Keypoint Detection by Handcrafted and Learned CNN Filters,ICCV 2019, [PDF], 深度学习特征点

• Self-Supervised 3D Keypoint Learning for Ego-motion Estimation,[PDF],[Youtube], 深度学习特征点

• VINS-Mono-Optimization, 实现点线紧耦合优化的VINS-Mono

• msckf_vio注释版本

• NetVLAD-pytorch, NetVLAD场景识别的pytorch实现

• High-Precision Localization Using Ground Texture (Micro-GPS),ECCV 2018,[PDF],[Project page],[code]，地向（摄像机朝向地面）SLAM，获得高精度重定位效果。

• PlaneSLAM, Paper: “On the Representation of Planes for Efficient Graph-based SLAM with High-level Features”

• XIVO: X Inertial-aided Visual Odometry and Sparse Mapping, an open-source repository for visual-inertial odometry/mapping.

• DeepTAM,ECCV 2018,[PDF],[Project page],a learnt system for keyframe-based dense camera tracking and mapping.

• iRotAvg, Why bundle adjust?,ICRA 2019,[PDF]

• Tactical Rewind: Self-Correction via Backtracking in Vision-and-Language Navigation,CVPR 2019,[PDF]，视觉+语言导航

• DOOR-SLAM

• An Evaluation of Feature Matchers for Fundamental Matrix Estimation,BMVC 2019,[PDF],[Project Page]，特征匹配

• A Tightly Coupled 3D Lidar and Inertial Odometry and Mapping Approach,ICRA 2019,[PDF],[Project Page]，紧耦合雷达+IMU SLAM

• On the Representation of Planes for Efficient Graph-based SLAM with High-level Features,利用平面信息的SLAM

• Visual Odometry Revisited: What Should Be Learnt?,arXiv 2019,[PDF], 深度学习深度+光流进行VO

• RF-Net: An End-to-End Image Matching Network based on Receptive Field,CVPR 2019,[PDF], 端到端图像匹配

• Fast-Planner,IEEE Robotics and Automation Letters (RA-L), 2019,[PDF], 无人机轨迹生成

• A general and flexible factor graph non-linear least square optimization framework,CoRR 2019,[PDF],[Project Page]

• Demo for Kalman filter in ranging system,卡尔曼滤波原理演示

• A Holistic Visual Place Recognition Approach using Lightweight CNNs for Severe ViewPoint and Appearance Changes，场景识别（外观与视角变化时）,训练和部署源码

• SIPs: Succinct Interest Points from Unsupervised Inlierness Probability Learning,3D Vision (3DV) 2019,[PDF]，RPG实验室出品，深度学习特征点（有特征描述子）

• Matching Features Without Descriptors: Implicitly Matched Interest Points,BMVC 2019,[PDF],RPG实验室出品，无需特征描述即可进行特征匹配

• Learning Lightweight Lane Detection CNNs by Self Attention Distillation (ICCV 2019),ICCV 2019,[PDF]，深度学习道路检测

• Awesome SLAM Datasets,史上最全SLAM数据集， 公众号说明: 最全 SLAM 开源数据集

• GNSS-INS-SIM,惯导融合模拟器，支持IMU数据，轨迹生成等

• Multi-Sensor Combined Navigation Program(GNSS, IMU, Camera and so on) 多源多传感器融合定位 GPS/INS组合导航

• SOSNet: Second Order Similarity Regularization for Local Descriptor Learning,CVPR 2019,[Project page] [Paper] [Poster] [Slides]，一种深度学习特征描述子

• Look No Deeper: Recognizing Places from Opposing Viewpoints under Varying Scene Appearance using Single-View Depth Estimation,ICRA 2019,[PDF],利用深度图像实现了大视角长时间的场景识别（根据深度图筛选得到不同深度层次的特征点然后与当前帧进行匹配，提高了场景召回率）

• CALC2.0,Convolutional Autoencoder for Loop Closure 2.0,用于闭环检测

• SegMap,RSS 2018,[PDF], 一种基于3D线段的地图表示，可用于场景识别/机器人定位/环境重建等

• MSCKF_VIO, a stereo version of MSCKF，基于MSCKF的双目VIO

• NetVLAD: CNN architecture for weakly supervised place recognition，CVPR 2016, CNN框架弱监督学习场景识别,[Project Page]

• easy_handeye,Simple, straighforward ROS library for hand-eye calibration

• SuperPoint-SLAM,利用SuperPoint替换ORB特征点

• PyRobot: An Open Source Robotics Research Platform

• From Coarse to Fine: Robust Hierarchical Localization at Large Scale with HF-Net,[PDF]

• Super fast implementation of ICP in CUDA

• A generic interface for disparity map and pointcloud insertion

• SPHORB: A Fast and Robust Binary Feature on the Sphere,International Journal of Computer Vision 2015,[PDF],[Project Page]

• BADSLAM: Bundle Adjusted Direct RGB-D SLAM,CVPR 2019,[PDF]

• High Speed and High Dynamic Range Video with an Event Camera,arXiv,[PDF],[Project Page]

• Awesome-VIO,Discuss about VIO in PaoPaoRobot group

• GyroAllan,陀螺仪随机误差的 Allan 方差分析, Another version

• Self-supervised Sparse-to-Dense: Self-supervised Depth Completion from LiDAR and Monocular Camera,ICRA 2019,[PDF], 优化LiDAR以及单目得到的深度图

• PlaneRCNN: 3D Plane Detection and Reconstruction from a Single Image,CVPR 2019,[PDF],[Project Page],通过单幅图像进行3D平面检测以及重建

• DBow3,注释版的DBow3代码

• Visual-Inertial Mapping with Non-Linear Factor Recovery,[PDF],[Project Page], 时空联合的VIO优化方案

• ICRA2019-paper-list,ICRA 2019论文列表（泡泡机器人出品暂时无链接）

• Fast Cylinder and Plane Extraction from Depth Cameras for Visual Odometry, IROS 2018,[PDF],利用深度图进行圆柱检测以及平面检测进行VO

• Solutions to assignments of Robot Mapping Course WS 2013/14 by Dr. Cyrill Stachniss at University of Freiburg,SLAM算法学习课后作业答案

• Direct sparse odometry combined with stereo cameras and IMU,双目DSO+IMU

• Direct Sparse Odometry with Stereo Cameras,双目DSO

• Python binding of SLAM graph optimization framework g2o,python版本的g2o实现

• SuperPoint: Self-Supervised Interest Point Detection and Description, CVPR 2018, [Paper], 深度学习描述子+描述

• ContextDesc: Local Descriptor Augmentation with Cross-Modality Context, CVPR 2019, [Paper], 深度学习描述子

• D2-Net: A Trainable CNN for Joint Description and Detection of Local Features, CVPR 2019, [Paper], [Project Page], 深度学习关键点+描述

• ROS interface for ORBSLAM2,ROS版本的ORBSLAM2

• CNN-SVO: Improving the Mapping in Semi-Direct Visual Odometry Using Single-Image Depth Prediction， [Paper]

• VINS-Mono-Learning，代码注释版VINS-Mono，初学者学习

• OpenVSLAM: Versatile Visual SLAM Framework, [Project Page]

• RESLAM: A real-time robust edge-based SLAM system, ICRA 2019, [Paper]

• PL-SLAM: a Stereo SLAM System through the Combination of Points and Line Segments, [Paper]，线特征SLAM

• Good Line Cutting: towards Accurate Pose Tracking of Line-assisted VO/VSLAM, ECCV 2018, [Project Page], 改进的PL-SLAM

• Spherical Regression: Learning Viewpoints, Surface Normals and 3D Rotations on n-Spheres, CVPR 2019, [Paper]

• svo_edgelet, 在线轨迹生成

• Drone SLAM project for Caltech's ME 134 Autonomy class, [PDF]

• Online Trajectory Generation of a MAV for Chasing a Moving Target in 3D Dense Environments, [Paper]

• PythonRobotics,[Paper], CppRobotics

• Bundle adjustment demo using Ceres Solver, [Blog], ceres实现BA

• CubeSLAM: Monocular 3D Object Detection and SLAM, [Paper]

• PointRCNN: 3D Object Proposal Generation and Detection from Point Cloud, CVPR 2019, [Paper]

• GIST-Global Image Descriptor, GIST描述子

• mav voxblox planning, MAV planning tools using voxblox as the map representation.

• Python Kalman Filter, 30行实现卡尔曼滤波

• vicalib, 视觉惯导系统标定工具

• BreezySLAM, 基于雷达的SLAM，支持Python(&Matlab, C++, and Java)

• Probabilistic-Robotics, 《概率机器人》中文版，书和课后习题

• Stanford Self Driving Car Code, [Paper], 斯坦福自动驾驶车代码

• Udacity Self-Driving Car Engineer Nanodegree projects

• Artificial Intelligence in Automotive Technology, TUM自动驾驶技术中的人工智能课程

• DeepMatchVO: Beyond Photometric Loss for Self-Supervised Ego-Motion Estimation,ICRA 2019, [Paper]

• GSLAM: A General SLAM Framework and Benchmark, CVPR 2019, [Paper], 集成了各种传感器输入的SLAM统一框架

• Visual-Odometric Localization and Mapping for Ground Vehicles Using SE(2)-XYZ Constraints，ICRA 2019,基于SE(2)-XYZ约束的VO系统

• Simple bag-of-words loop closure for visual SLAM, [Blog], 回环

• FBOW (Fast Bag of Words), an extremmely optimized version of the DBow2/DBow3 libraries,优化版本的DBow2/DBow3

• Multi-State Constraint Kalman Filter (MSCKF) for Vision-aided Inertial Navigation(master's thesis)

• MSCKF, MSCKF中文注释版

• Calibration algorithm for a camera odometry system, VO系统的标定程序

• Modified version of VINS-Mono, 注释版本VINS Mono

• Extreme Relative Pose Estimation for RGB-D Scans via Scene Completion,[Paper]

• Implementation of EPnP algorithm with Eigen,利用Eigen编写的EPnP

• Real-time SLAM system with deep features, 深度学习描述子(ORB vs. GCNv2)

• Unsupervised Learning of Monocular Depth Estimation and Visual Odometry with Deep Feature Reconstruction, CVPR 2018, 无监督单目深度恢复以及VO

• ORB-SLAM-windows, Windows版本的ORB-SLAM

• StructVIO : Visual-inertial Odometry with Structural Regularity of Man-made Environments,[Project Page]

• KalmanFiltering, 各种卡尔曼滤波器的demo

• Stereo Odometry based on careful Feature selection and Tracking, [Paper], C++ OpenCV实现SOFT

• Visual SLAM with RGB-D Cameras based on Pose Graph Optimization

• Multi-threaded generic RANSAC implemetation, 多线程RANSAC

• Visual Odometry with Drift-Free Rotation Estimation Using Indoor Scene Regularities, BMVC 2017, [Project Page]，利用平面正交信息进行VO

• ICE-BA, CVPR 2018, [Paper]

• GraphSfM: Robust and Efficient Graph-based Structure from Motion, [Project Page]

• LOAM_NOTED, loam中文注解版

• Divide and Conquer: Effcient Density-Based Tracking of 3D Sensors in Manhattan Worlds,ACCV 2016,[Project Page],曼哈顿世界利用深度传感器进行旋转量平移量分离优化

• Real-time Manhattan World Rotation Estimation in 3D,IROS 2015,实时曼哈顿世界旋转估计

• Event-based Vision Resources，关于事件相机的资源

• AutonomousVehiclePaper，无人驾驶相关论文速递

• Segmentation.X, Segmentation相关论文&代码

• CVPR-2019, CVPR 2019 论文开源项目合集

• awesome-slam, SLAM合集

• awesome-visual-slam, 视觉SLAM合集

• Papers with code, 周更论文with代码

• Awesome Human Pose Estimation,awesome-object-pose, 位姿估计合集

• MVision, 大礼包：机器人视觉 移动机器人 VS-SLAM ORB-SLAM2 深度学习目标检测 yolov3 行为检测 opencv PCL 机器学习 无人驾驶

Pose/Object tracking

• Unsupervised person re-identification by soft multilabel learning,CVPR 2019, [Paper]

• FCOS: Fully Convolutional One-Stage Object Detection,ICCV 2019, [Paper]

• Hand Detection and Orientation Estimation

• Spatial-Temporal Person Re-identification,AAAI 2019,[Paper]

• A tiny, friendly, strong pytorch implement of person re-identification baseline. Tutorial,CVPR 2019, [Paper]

• Progressive Pose Attention for Person Image Generation,CVPR 2019,[Paper]

• FSA-Net: Learning Fine-Grained Structure Aggregation for Head Pose Estimation from a Single Image, CVPR 2019,[Paper]

• An unoffical implemention for paper "Fast Human Pose Estimation", CVPR 2019,[Paper]

• Real-time single person pose estimation for Android and iOS,手机端实现人体位姿估计

• Basics of 2D and 3D Human Pose Estimation,人体姿态估计入门

• Libra R-CNN: Towards Balanced Learning for Object Detection

• High-resolution networks (HRNets) for object detection, [Paper]

• Learning Correspondence from the Cycle-Consistency of Time, CVPR 2019, [Paper]

• PVNet: Pixel-wise Voting Network for 6DoF Pose Estimation, CVPR 2019, [Paper], [Project Page]

• Self-Supervised Learning of 3D Human Pose using Multi-view Geometry, CVPR 2018, [Paper]

• PifPaf: Composite Fields for Human Pose Estimation, [Paper]

• Deep High-Resolution Representation Learning for Human Pose Estimation,CVPR 2019, [Paper], [Project Page]

• PoseFlow: Efficient Online Pose Tracking), BMVC 2018, [Paper]

• A Bottom-Up Clustering Approach to Unsupervised Person Re-identification，AAAI 2019, 重定位

• Fast Online Object Tracking and Segmentation: A Unifying Approach,CVPR 2019,[Paper] [Video] [Project Page]

• SimpleDet - A Simple and Versatile Framework for Object Detection and Instance Recognition,[Paper]

Depth/Disparity & Flow estimation

• [Depth]SemiGlobalMatching, SGM双目立体匹配算法完整实现，代码规范，注释丰富且清晰，CSDN同步教学

• PointMVSNet: Point-based Multi-view Stereo Network,ICCV 2019,[Paper]

• DeepLiDAR,CVPR 2019, [Paper], 单张RGB图像+稀疏雷达数据进行室外场景深度估计

• Real-Time Monocular Depth Estimation using Synthetic Data with Domain Adaptation via Image Style Transfer,CVPR 2018, [Paper]

• Learning Single-Image Depth from Videos using Quality Assessment Networks,CVPR 2019, [Paper], [Project Page]

• SCDA: Adapting Object Detectors via Selective Cross-Domain Alignment,CVPR 2019, [Paper], [Project Page]

• Learning monocular depth estimation infusing traditional stereo knowledge,CVPR 2019,[PDF]

• HPLFlowNet: Hierarchical Permutohedral Lattice FlowNet for Scene Flow Estimation on Large-scale Point Clouds,CVPR 2019,[Paper]

• GA-Net: Guided Aggregation Net for End-to-end Stereo Matching,CVPR 2019,[Paper]

• DPSNet: End-to-end Deep Plane Sweep Stereo,ICLR 2019,[Paper]

• Fast Depth Densification for Occlusion-aware Augmented Reality, SIGGRAPH-Asia 2018, [Project Page],another version

• Learning To Adapt For Stereo, CVPR 2019, [Paper]

• Pyramid Stereo Matching Network,[Paper]

• Bridging Stereo Matching and Optical Flow via Spatiotemporal Correspondence, [Paper]

• Sparse Depth Completion, [Paper], RGB图像辅助雷达深度估计

• GASDA, CVPR 2019, [Paper]

• MVSNet: Depth Inference for Unstructured Multi-view Stereo, [Paper], 非官方实现版本的MVSNet

• Stereo R-CNN based 3D Object Detection for Autonomous Driving, CVPR 2019, [Paper]

• Real-time self-adaptive deep stereo, CVPR 2019, [Paper]

• High Quality Monocular Depth Estimation via Transfer Learning,CVPR 2019, [Paper], [Project Page]

• Group-wise Correlation Stereo Network,CVPR 2019, [Paper]

• DeepMVS: Learning Multi-View Stereopsis, CVPR 2018,[Project Page],多目深度估计

• FlowNet 2.0: Evolution of Optical Flow Estimation with Deep Networks, CVPR 2017, 深度学习光流恢复

• StereoVision-ADCensus,深度恢复代码集合(ADCensus, SGBM, BM)

• SegStereo: Exploiting Semantic Information for Disparity Estimation, 探究语义信息在深度估计中的作用

• Light Filed Depth Estimation using GAN，利用GAN进行光场深度恢复

• EV-FlowNet: Self-Supervised Optical Flow for Event-based Cameras,Proceedings of Robotics 2018,[Paper]

• DF-Net: Unsupervised Joint Learning of Depth and Flow using Cross-Task Consistency, ECCV 2018, [Paper]

• GeoNet: Unsupervised Learning of Dense Depth, Optical Flow and Camera Pose, CVPR 2018, [Paper]

3D & Graphic

• PRNet: Self-Supervised Learning for Partial-to-Partial Registration,NeurIPS 2019

• Learning to Reconstruct 3D Human Pose and Shape via Model-fitting in the Loop,ICCV 2019, [Paper] , [Project Page]

• Cross View Fusion for 3D Human Pose Estimation,ICCV 2019, [Paper] ,跨视角3D位姿估计

• MVF-Net: Multi-View 3D Face Morphable Model Regression,多视角3D人脸重建, [Paper]

• KillingFusion

• ReFusion: 3D Reconstruction in Dynamic Environments for RGB-D Cameras Exploiting Residuals, [Paper]

• densebody_pytorch, [Paper]

• Single-Image Piece-wise Planar 3D Reconstruction via Associative Embedding,CVPR 2019, [Paper], 单目3D重建

• HorizonNet: Learning Room Layout with 1D Representation and Pano Stretch Data Augmentation,CVPR 2019, [Paper], 深度学习全景转3D

• Adaptive O-CNN: A Patch-based Deep Representation of 3D Shapes,SIGGRAPH Asia 2018, [Project Page]

Other Collections

• Matrix-Calculus, 矩阵求导方法

• Mathematics,数学知识点滴积累,矩阵,数值优化,神经网络反向传播,图优化,概率论,随机过程,卡尔曼滤波,粒子滤波,数学函数拟合

• chinese-independent-blogs, 中文独立博客集锦

• StructureFlow: Image Inpainting via Structure-aware Appearance Flow,图像inpainting

• free-books,互联网上的免费书籍

• AcademicPages,通用的学术主页模版

• MMdnn,实现深度学习模型之间的相互转换

• tensorflow2caffemodel,tensorflow模型转caffemodel

• lihang-code,《统计学习方法》的代码实现

• sse2neon,sse2neon,SSE转neon，嵌入式移植时可能会用到;

• Production-Level-Deep-Learning,深度学习模型部署流程

• 动手学深度学习Dive-into-DL-PyTorch

• machine-learning-yearning-cn，Machine Learning Yearning 中文版 - 《机器学习训练秘籍》 - Andrew Ng 著

• academicpages.github.io，学术主页模板

• Coursera-ML-AndrewNg-Notes,吴恩达老师的机器学习课程个人笔记

• machine-learning-notes,机器学习，概率模型和深度学习的讲义(1500+页)和视频链接

• CNN-Visualization,CNN可视化、理解CNN

• Awesome Semantic Segmentation, 语义分割集合

• IROS2018 SLAM Collections, IROS 2018集合

• VP-SLAM-SC-papers,Visual Positioning & SLAM & Spatial Cognition 论文统计与分析

• Awesome System for Machine Learning

• Machine-Learning-With-Python, 《机器学习实战》python代码实现

• How to learn robotics, 开源机器人学学习指南

• Awesome Deep Vision,DL在CV领域的应用

• Single-Image-Super-Resolution, 一个有关图像超分辨的合集

• ai report, AI相关的研究报告

• State-of-the-art papers and code,搜集了目前sota的论文以及代码

• CVPR 2019 (Papers/Codes/Project/Paper reading)

• A curated list of papers & resources linked to 3D reconstruction from images,有关三维重建的论文汇总

• SLAM-Jobs, SLAM/SFM求职指南

• Spatial Attentive Single-Image Deraining with a High Quality Real Rain Dataset,CVPR 2019,去雨

• Densely Connected Pyramid Dehazing Network,CVPR 2018,去雾

• MMSR，MMLAB推出的超分辨工具箱

• 深度学习OCR

• 西瓜书🍉学习笔记

• awesome-reinforcement-learning-zh,强化学习从入门到放弃的资料

• Deep Plug-and-Play Super-Resolution for Arbitrary Blur Kernels,CVPR 2019,超分辨

• Cool Fashion Papers, Cool resources about Fashion + AI.

• Deep Flow-Guided Video Inpainting,CVPR 2019, [Paper] ,图像修复

• YOLACT: Real-time Instance Segmentation

• LPRNet: License Plate Recognition via Deep Neural Networks, [Paper]

• CHINESE-OCR, 运用tf实现自然场景文字检测

• BeautyCamera, 美颜相机，具有人脸检测、磨皮美白人脸、滤镜、调节图片、摄像功能

• CV-arXiv-Daily, 分享计算机视觉每天的arXiv文章

• Pluralistic-Inpainting, ArXiv | Project Page | Online Demo | Video(demo)

• An Interactive Introduction to Fourier Transforms, 超棒的傅里叶变换图形化解释

• pumpkin-book, 《机器学习》（西瓜书）公式推导解析

• Julia

• A Julia machine learning framework，一种基于Julia的机器学习框架

• High-Performance Face Recognition Library on PyTorch，人脸识别库

• Deep-Learning-Coursera，深度学习教程（deeplearning.ai）

• The best resources around Machine Learning

• VGGFace2: A dataset for recognising faces across pose and age

• Statistical learning methods，统计学习方法

• End-to-end Adversarial Learning for Generative Conversational Agents，2017，介绍了一种端到端的基于GAN的聊天机器人

• Residual Non-local Attention Networks for Image Restoration,ICLR 2019.

• MSGAN: Mode Seeking Generative Adversarial Networks for Diverse Image Synthesis, CVPR 2019,[Paper]

• SPADE: Semantic Image Synthesis with Spatially-Adaptive Normalization,CVPR 2019, [Project Page]

• Faceswap with Pytorch or DeepFake with Pytorch, 换脸

• DeepFaceLab, 换脸

Contribute

❤ Please feel free to pull requests to add links.

License

Star History