3D Object Detection On Sun Rgbd Val

评估指标

mAP@0.25

mAP@0.5

评测结果

各个模型在此基准测试上的表现结果

模型名称	mAP@0.25	mAP@0.5	Paper Title	Repository
V-DETR	68.0	51.1	V-DETR: DETR with Vertex Relative Position Encoding for 3D Object Detection
Frustum VoxNet v2 (FPN + 3D ResNetFCN6 V2)	45.0	-	3D Object Detection and Instance Segmentation from 3D Range and 2D Color Images	-
COG+surface	51.0	-	3D Object Detection With Latent Support Surfaces	-
DEST (based on V-DETR) (TTA)	69.2	52.2	-	-
CAGroup3D(Geo only)	66.8	50.2	CAGroup3D: Class-Aware Grouping for 3D Object Detection on Point Clouds
BRNet(Geo only)	61.1	43.7	Back-tracing Representative Points for Voting-based 3D Object Detection in Point Clouds
DEST (based on GroupFree3D)	65.3	48.4	-	-
HGNet (Geo only)	61.6	-	A Hierarchical Graph Network for 3D Object Detection on Point Clouds	-
RBGNet(Geo only)	64.1	47.2	RBGNet: Ray-based Grouping for 3D Object Detection
COG+surface+context	54.3	-	Clouds of Oriented Gradients for 3D Detection of Objects, Surfaces, and Indoor Scene Layouts	-
Frustum VoxNet (+3D ResNetFCN6)	37.7	-	Frustum VoxNet for 3D object detection from RGB-D or Depth images	-
H3DNet	60.1	39.0	H3DNet: 3D Object Detection Using Hybrid Geometric Primitives
VoteNet (PC-FractalDB)	60.2	35.2	Point Cloud Pre-Training With Natural 3D Structures
TR3D (Geo only)	67.1	50.4	TR3D: Towards Real-Time Indoor 3D Object Detection
2D-driven	45.1	-	2D-Driven 3D Object Detection in RGB-D Images	-
GroupFree3D(Geo only)	63.0	45.2	Group-Free 3D Object Detection via Transformers
Frustum VoxNet (YOLO v3+3D ResNetFCN6)	-	-	Frustum VoxNet for 3D object detection from RGB-D or Depth images	-
TR3D+FF	69.4	53.4	TR3D: Towards Real-Time Indoor 3D Object Detection
Point-GCC+TR3D+FF	69.7	54.0	Point-GCC: Universal Self-supervised 3D Scene Pre-training via Geometry-Color Contrast
DeMF	67.4	51.2	Boosting 3D Object Detection via Object-Focused Image Fusion

0 of 32 row(s) selected.