- 博客(42)
- 收藏
- 关注
原创 (2022 IV) RCBEV
这篇是清华大学发表在IV(IEEE Transactions on Intelligent Vehicles)上的文章: Bridging the view disparity between radar and camera features for multi-modal fusion 3d object detection文章信息讲得比较细致,非常值得一看;
2023-05-03 17:20:18
980
1
原创 CRN Camera Radar Net for Accurate, Robust, Efficient 3D Perception
这是发表在ICLR2023 workshop on Scene Representations for Autonomous Driving上的文章,目前在nuScenes的camera+Radar的3D目标检测赛道上刷到了第一,非常值得一看。
2023-04-30 22:23:48
781
原创 CRAFT: Camera-Radar 3D Object Detection with Spatio-Contextual Fusion Transformer
radar, camera fusion, aaai2022
2023-04-09 18:19:48
571
原创 MVFusion: Multi-View 3D Object Detection with Semantic-aligned Radar and Camera Fusion
MVFusion, 多视图radar+camera融合的3D目标检测, nuScenes上SOTA, ICRA2023.
2023-04-08 18:05:05
326
原创 Towards Deep Radar Perception for Autonomous Driving: Datasets, Methods, and Challenges
大家直接看原文把
2023-03-26 18:10:52
162
原创 (2019.01, iclr) Decoupled Weight Decay Regularization
讲adamW的论文, 除了纯SGD, L2 != weight_decay
2023-02-28 19:57:43
324
1
原创 (2018, sensors) SECOND: Sparsely Embedded Convolutional Detection
second论文阅读记录
2023-02-24 20:49:05
151
原创 (2021.06, iccv) Swin Transformer Hierarchical Vision Transformer using Shifted Windows
swin transfomer!!!
2023-02-23 21:30:48
160
原创 Cross-view Transformers for real-time Map-view Semantic Segmentation
Cross-view Transformers for real-time Map-view Semantic Segmentation
2023-02-19 19:29:48
637
1
原创 (2021.06, cvpr) Center-Based 3D Object Detection and Tracking
3d 版本center net
2023-02-17 00:22:59
158
原创 Monocular Localization in HD Maps by Combining Semantic Segmentation and Distance Transform
高精地图+图片语义分割定位,用“光度误差”代替明确的数据关联过程
2023-02-06 15:19:24
355
原创 (2022.05) BEVFusion: Multi-Task Multi-Sensor Fusion with Unified Bird‘s-Eye View Representation
mit 版本的BEVFusion
2023-02-04 23:07:40
466
原创 (2021, icra) Visual Semantic Localization based on HD Map for Autonomous Vehicles in Urban Scenarios
视觉+高精地图定位
2023-02-03 17:47:10
501
原创 (2022, iv) A Survey on Map-Based Localization Techniques for Autonomous Vehicles
基于地图的定位方案综述
2023-02-02 16:08:56
191
原创 Coarse-to-fine Semantic Localization with HD Map for Autonomous Driving in Structural Scenes
由粗到精,基于语义信息和高精地图定位
2023-01-17 16:55:25
275
原创 (2020.01) Monocular localization with Vector HD Map (MLVHM): A Low-Cost Method for Commercial IVs
单目加上高精地图定位
2023-01-16 17:57:01
233
原创 论文阅读笔记:(2022) Delving into the Devils of Bird‘s-eye-view Perception: A Review, Evaluation and Recipe
bev 感知review
2022-11-21 15:49:18
963
1
原创 论文阅读笔记: (2022 TPAMI) Ultra Fast Deep Lane Detection with Hybrid Anchor Driven Ordinal Classification
车道线检测是自动驾驶中的基础任务之一,我们今天来看一篇来自浙大的TPAMI 2022的文章!
2022-09-12 18:01:41
1146
原创 论文阅读笔记:(2014) Effect of field of view on the accuracy of camera calibration
长焦(小FOV)的相机标定
2022-08-21 19:10:04
723
原创 论文阅读笔记:(2022.05) Observability-Aware Intrinsic and Extrinsic Calibration of LiDAR-IMU Systems
介绍激光雷达和IMU的内外参标定
2022-08-20 17:43:30
715
1
原创 论文阅读笔记:(1998) A flexible new technique for camera calibration
大名鼎鼎的相机内参的张氏标定法,我们今天来一探究竟!
2022-07-24 17:33:43
949
原创 论文阅读笔记: (2022.05, icra) Traffic Context Aware Data Augmentation for Rare Object Detection in Autonom
这篇文章是nullmax发到icra上的, 主要讲了用domain迁移的方法制作稀有场景的数据集
2022-07-19 16:49:20
257
原创 论文阅读笔记:(2021)MetroLoc Metro Vehicle Mapping and Localization with LiDAR-Camera-Inertial Integration
一、创新点,主要贡献1. 一个地铁场景的实时slam 框架2. 以IMU为中心的激光-视觉-惯性的紧耦合(因子图)3. 高效,可靠的Lidar特征提取框架4. 实验充分二、精度短距离:Experiment-1: 特征丰富的站台和岔路口, 450mExperiment-2:隧道, 750m长距离, 地下四站:速度分析:i9, 64GB, 多线程实时:三、实现...
2022-05-30 15:07:40
230
2
原创 论文阅读笔记:(2015, ijrr) Keyframe-based visual–inertial odometry using nonlinear optimization
算是基于滑窗的VIO的必读文章吧,很详细地说了边缘化~paper:http://in.ruc.edu.cn/wp-content/uploads/2021/01/Keyframe-Based-Visual-Inertial-Odometry-Using-Nonlinear-Optimization.pdfhttp://in.ruc.edu.cn/wp-content/uploads/2021/01/Keyframe-Based-Visual-Inertial-Odometry-Using-Nonl.
2022-05-22 22:08:21
512
原创 论文阅读笔记: (2018.08, iros) LDSO: Direct Sparse Odometry with Loop Closure
我自己关于DSO相关的论文阅读笔记汇总在这里:论文阅读笔记:vslam之DSO系列_chaoqinyou的博客-CSDN博客想用这篇文章,把DSO系列的文章穿起来读一下!DSO系列文章是稀疏直接法vslam的代表,这个系列的文章大多发表在机器人或者视觉的顶刊或者顶刊上,也有不少开源的实现,非常值得学习!TUM Computer Vision Group的链接镇场:Computer Vision Group - Visual SLAMVisual SLAMVisual SLAMIn Simultaneou
2022-05-02 11:43:21
697
原创 论文阅读笔记: (2020.01 cvpr) ImVoteNet: Boosting 3D Object Detection in Point Clouds with Image Votes
paper:https://arxiv.org/abs/2001.10692https://arxiv.org/abs/2001.10692官方实现:GitHub - saic-vul/imvoxelnet: [WACV2022] ImVoxelNet: Image to Voxels Projection for Monocular and Multi-View General-Purpose 3D Object Detectionhttps://github.com/saic-vul/imvox
2022-04-24 16:22:31
3806
原创 docker相关: 安装, docker file等
docker guide:Docker overview | Docker Documentationhttps://docs.docker.com/get-started/overview/nvidia-docker:Installation Guide — NVIDIA Cloud Native Technologies documentationhttps://docs.nvidia.com/datacenter/cloud-native/container-toolkit/install
2022-04-23 11:30:59
201
原创 论文阅读笔记:(2021.10 CoRL) DETR3D: 3D Object Detection from Multi-view Images via 3D-to-2D Queries
论文地址:DETR3D: 3D Object Detection from Multi-view Images via 3D-to-2D Queries | OpenReviewWe introduce a framework for multi-camera 3D object detection. In contrast to existing works, which estimate 3D bounding boxes directly from monocular images or use de
2022-04-19 22:12:20
694
2
原创 论文阅读笔记:vslam之DSO系列
想用这篇文章,把DSO系列的文章穿起来读一下!DSO系列文章是稀疏直接法vslam的代表,这个系列的文章大多发表在机器人或者视觉的顶刊或者顶刊上,也有不少开源的实现,非常值得学习!TUM Computer Vision Group的链接镇场:Computer Vision Group - Visual SLAMVisual SLAMVisual SLAMIn Simultaneous Localization And Mapping, we track the pose of the sensor.
2022-04-09 17:49:15
2168
2
原创 非线性最小二乘, BA(Bundle Adjustment)
在看高翔博士的《视觉slam十四讲》,把看到的BA相关的东西记录一下,方便自己之后复习,有理解不对的地方,还请指正!一、非线性最小二乘1.1 总体思路, 一阶法和二阶法最小二乘问题 ,我们称F(x)是代价函数, f(x)是误差函数;直接求再比较各个极值点/鞍点的大小一般是比较难的。因此, 一般采取迭代的方法: 即从某个初值x0开始,每一次寻找一个可以使得F(x)变小的增量delta_x, 当增量delta_x比较小的时候,我们就认为认为达到了极值。一阶法寻找delta_x: 将F(x).
2022-04-09 12:17:37
4301
原创 二分图匹配,Hangarian算法, Kuhn-Munkres算法
1. 比较好的文章:图论总结(一)二分图最大匹配_LinHunYoR的博客-CSDN博客_二分图最大匹配https://blog.csdn.net/qq_38956769/article/details/80238896https://brc2.com/the-algorithm-workshop/https://brc2.com/the-algorithm-workshop/1. 二分图最优匹配:匹配:在二分图G的一个子图M中, M的边集{E}中的任意两条边都不交汇于同一个结点,则称M是一个匹
2022-03-25 09:23:23
1199
原创 论文阅读笔记:(2021.06 iccv) Geometry-based Distance Decomposition for Monocular 3D Object Detection
好思路:不直接regression深度,而是regression 3D框高度在图片上的投影,再用相似三角形算出深度;paper:http://openaccess.thecvf.com//content/ICCV2021/papers/Shi_Geometry-Based_Distance_Decomposition_for_Monocular_3D_Object_Detection_ICCV_2021_paper.pdfhttp://openaccess.thecvf.com//content/ICC
2022-02-07 18:02:35
2730
原创 论文阅读笔记:(2021 cvpr) Depth-conditioned Dynamic Message Propagation for Monocular 3D Object Detection
paper:CVPR 2021 Open Access Repositoryhttps://openaccess.thecvf.com/content/CVPR2021/html/Wang_Depth-Conditioned_Dynamic_Message_Propagation_for_Monocular_3D_Object_Detection_CVPR_2021_paper.htmlcode:GitHub - fudan-zvg/DDMP: [CVPR 2021] Depth-conditioned
2022-02-07 17:55:08
2160
原创 论文阅读笔记:(2021.06 cvpr) Categorical Depth Distribution Network for Monocular 3D Object Detection
paper:https://openaccess.thecvf.com/content/CVPR2021/papers/Reading_Categorical_Depth_Distribution_Network_for_Monocular_3D_Object_Detection_CVPR_2021_paper.pdfhttps://openaccess.thecvf.com/content/CVPR2021/papers/Reading_Categorical_Depth_Distribution_Net
2022-02-07 16:55:25
822
原创 论文阅读笔记:(2021.06, cvpr) Monocular 3D Object Detection: An Extrinsic Parameter Free Approach
这是一篇发表在了cvpr 2021上,能够在线估计外参的单目3D目标检测算法。paper:https://openaccess.thecvf.com/content/CVPR2021/papers/Zhou_Monocular_3D_Object_Detection_An_Extrinsic_Parameter_Free_Approach_CVPR_2021_paper.pdfhttps://openaccess.thecvf.com/content/CVPR2021/papers/Zhou_Mon
2022-02-04 17:17:07
2952
原创 论文阅读笔记:(2021.06 cvpr) Objects are Different: Flexible Monocular 3D Object Detection
本文介绍cvpr 2021的MonoFLEX,论文的着眼点是解决遮挡/截断物体的3D检测。这个方法也是目前(截止2022.01)没有extra traning data的条件下, kitti-car-moderate上结果最好的方法。block-diagram有点复杂,不过挺值得一看的!code: https://github.com/zhangyp15/MonoFlexhttps://github.com/zhangyp15/MonoFlexpaper:http://arxiv.org/.
2022-01-30 11:06:04
2425
原创 论文阅读笔记:(2021.08, iccv) Is Pseudo-Lidar needed for Monocular 3D Object detection?
DD3D发表在了iccv 2021上,是目前(2022.01)性能最好的单目3D目标检测模型之一,值得一看。论文:ICCV 2021 Open Access Repositoryhttps://openaccess.thecvf.com/content/ICCV2021/html/Park_Is_Pseudo-Lidar_Needed_for_Monocular_3D_Object_Detection_ICCV_2021_paper.html代码:https://github.com/TRI-ML/
2022-01-29 16:41:03
2428
空空如也
空空如也
TA创建的收藏夹 TA关注的收藏夹
TA关注的人