一、关注的论文类别
- SLAM & VO & Tracking
- 机器智能系统性能优化
- 语义
- 传感器相关
- 地图和重建
- 数据集相关
- 机器智能系统的崭新研究
- 数学计算技术
- 人机协作 & 多机协作
- 移动机器人 & 无人车
二、论文关键信息
0. SLAM & VO & Tracking
论文关键信息 | 论文题目 |
---|---|
VINS综述 | A review of vio from filtering and optimisation perspectives |
Rolling Shutter,李明扬 | Real-time Motion Tracking on a Cellphone using Inertial Sensing and a Rolling-Shutter Camera |
Rolling Shutter | Monocular Visual Odometry with a Rolling Shutter Camera |
点线VO,ICRA2017,应对低纹理、光照变化和快速移动;跟踪部分维持关键帧深度图,光度误差+边缘几何误差 | Direct Monocular Odometry Using Points and Lines |
点线SLAM,应对低纹理(城市,室内),基于ORBSLAM | PL-SLAM: Real-Time Monocular Visual SLAM with Points and Lines |
只关注兴趣区域的DSO,场景解析辅助特征选取 | SalientDSO: Bringing Attention to Direct Sparse Odometry |
拓扑SLAM,Voronoi图 | Topological Simultaneous Localization and Mapping (SLAM): Toward Exact Localization Without Explicit Localization |
Cloud Based Localization | Global Localization from Monocular SLAM on a Mobile Phone |
医疗SLAM | Markerless Inside-Out Tracking for Interventional Applications |
DVS,事件相机,1MHz,无结构测量模型 | Event-Based Visual Inertial Odometry |
结合语义,适用平面低纹理场景,Shichao Yang, Michael Kaess | Pop-up SLAM: Semantic Monocular Plane SLAM for Low-texture Environments |
关键点跟踪,考虑形变和光照,Stefano Soatto | Real-time feature tracking and outlier rejection with changes in illumination |
区域跟踪,考虑形变和光照,改进的SSD | Real-Time Tracking of Image Regions with Changes in Geometry and Illumination |
平面物体跟踪,考虑光照,利用梯度朝向信息 | Illumination Insensitive Efficient Second-order Minimization for Planar Object Tracking |
1. 机器智能系统性能优化
论文关键信息 | 论文题目 |
---|---|
网络压缩 | LCDet: Low-Complexity Fully-Convolutional Neural Networks for Object Detection in Embedded Systems |
提高图像分类精度 | Dilated Residual Networks |
神经网络效率精度优化 | Model Complexity-Accuracy Trade-off for a Convolutional Neural Network |
2. 语义
论文关键信息 | 论文题目 |
---|---|
快速语义三维重建 | Fast and Accurate Semantic Mapping through Geometric-based Incremental Segmentation |
语义视觉定位,描述子学习,Andreas Geiger | Semantic Visual Localization |
语义SLAM,目标检测框作为landmark | Dual Quadrics from Object Detection Bounding Boxes as Landmark Representations in SLAM |
3. 传感器相关
论文关键信息 | 论文题目 |
---|---|
多传感器标定,带源码 | LiDAR-Camera Calibration using 3D-3D Point correspondences |
多传感器标定,激光雷达和三目相机,带源码 | Automatic Extrinsic Calibration for Lidar-Stereo Vehicle Sensor Setups |
传感器网络,综述 | A Review of Localization and Tracking Algorithms in Wireless Sensor Networks |
DVS,事件相机,1MHz,无结构测量模型 | Event-Based Visual Inertial Odometry |
深度相机 | Intel RealSense Stereoscopic Depth Cameras |
4. 地图和重建
论文关键信息 | 论文题目 |
---|---|
在OKVIS基础上添加闭环检测和重定位的工作 | Keyframe-Based Visual-Inertial Online SLAM with Relocalization |
ETHZ,ROVIO,VIO,多阶段建图,地图复用,开源代码,ROS | maplab: An Open Framework for Research in Visual-inertial Mapping and Localization |
跳表树,时间效率高于Octomap | SkiMap: An Efficient Mapping Framework for Robot Navigation |
todo | MapNet: Geometry-Aware Learning of Maps for Camera Localization |
点云处理,模板,支持任意维度,少外部依赖,开源 | cilantro: a lean, versatile, and efficient library for point cloud data processing |
5. 数据集相关
论文关键信息 | 论文题目 |
---|---|
数据集质量评估,CV-HAZOP清单 | Analyzing Computer Vision Data - The Good, the Bad and the Ugly |
数据集,三维重建,标注(相机位姿/平面重建/语义分割),1513个场景,250万张图,RGB-D采集系统 | ScanNet: Richly-annotated 3D Reconstructions of Indoor Scenes |
数据集,遥感图像,OSM,武汉大学 | RSI-CB: A Large Scale Remote Sensing Image Classification Benchmark via Crowdsource Data |
数据集,视频,80个AVA,人类行为理解 | AVA: A Video Dataset of Spatio-temporally Localized Atomic Visual Actions |
数据集,目标识别,Lifelong | CoRe50: a New Dataset and Benchmark for Continuous Object Recognition |
6. 机器智能系统的崭新研究
论文关键信息 | 论文题目 |
---|---|
将GUI截屏转为代码,end-to-end | pix2code: Generating Code from a Graphical User Interface Screenshot |
中国诗词生成,注意力机制,记忆增强,提升美学性 | Flexible and Creative Chinese Poetry Generation Using Neural Memory |
司机注视点预测,ADAS,DR(eye)VE数据集 | Predicting the Driver’s Focus of Attention: the DR(eye)VE Project |
图灵测试不适用于评估机器智能,标准化测试 | A Survey of Question Answering for Math and Science Problem |
7. 数学计算技术
论文关键信息 | 论文题目 |
---|---|
矩阵加速 | CLBlast: A Tuned OpenCL BLAS Library |
8. 人机协作&多机协作
论文关键信息 | 论文题目 |
---|---|
人人交互启发人机交互 | Human-Robot Collaboration - From Psychology to Social Robotics |
多机器人系统,快速部署,廉价硬件 | Swarm-Enabling Technology for Multi-Robot Systems |
拥挤环境导航,合作机制建模,轨迹预测 | Modeling Cooperative Navigation in Dense Human Crowds |
人机协作,以统一的框架优化人、机、环境系统的最佳协作 | A Mathematical Theory of Human Machine Teaming |
9. 移动机器人 & 无人车
论文关键信息 | 论文题目 |
---|---|
避障,双目相机,深度图划分子区域,模糊逻辑控制器 | Obstacle Avoidance Using Stereo Camera |
其他
非线性最小二乘建模 Nonlinear State Estimation for Robotics and Computer Vision Applications: An Overview Arun Das
视觉场景识别,全图描述子 Efficient Decentralized Visual Place Recognition From Full-Image Descriptors
仿真器,物理引擎,微软,带源码 AirSim - High-Fidelity Visual and Physical Simulation for Autonomous Vehicles
迁移学习,综述 A Cross-Dataset Recognition: A Survey
实时稠密重建,深度滤波 Video-based, real-time multi-view stereo
单目稠密重建,深度滤波,开源,REMODE,CUDA实现30FPS,Matia Pizzoli, Christian Forster and Davide Scaramuzza REMODE: Probabilistic, Monocular Dense Reconstruction in Real Time
单目深度图估计和运动估计,无监督,Google,David Lowe,开源,SfMLearner Unsupervised Learning of Depth and Ego-Motion from Video
长距离车辆高精度路径跟踪,局部平面假设先验,单目视觉尺度修正,Tim Barfoot Monocular Visual Teach and Repeat Aided by Local Ground Planarity
单目,3D人体姿态估计,迁移学习 Monocular 3D Human Pose Estimation In The Wild Using Improved CNN Supervision
与Peter Abbeel合作,人类行动模式启发,认知计算模型 Robot Planning with Mathematical Models of Human State and Action