Method

PUDet_Lidar [PUDet]


Submitted on 5 Apr. 2024 09:10 by
yue Zhang (Tsinghua University)

Running time:0.3 s
Environment:GPU @ 2.5 Ghz (Python)

Method Description:
Using Point Upsampling method, make the sparse
foreground point cloud have more context
information, which can help the 3D Object Detection.
Parameters:
None
Latex Bibtex:
None

Detailed Results

Object detection and orientation estimation results. Results for object detection are given in terms of average precision (AP) and results for joint object detection and orientation estimation are provided in terms of average orientation similarity (AOS).


Benchmark Easy Moderate Hard
Car (Detection) 96.32 % 92.77 % 90.02 %
Car (Orientation) 96.31 % 92.62 % 89.81 %
Car (3D Detection) 87.85 % 79.34 % 74.58 %
Car (Bird's Eye View) 92.68 % 88.42 % 83.70 %
Pedestrian (Detection) 62.39 % 51.41 % 49.08 %
Pedestrian (Orientation) 58.44 % 47.30 % 44.83 %
Pedestrian (3D Detection) 46.00 % 37.42 % 35.12 %
Pedestrian (Bird's Eye View) 50.24 % 41.48 % 39.22 %
Cyclist (Detection) 79.83 % 65.72 % 59.32 %
Cyclist (Orientation) 79.42 % 65.35 % 58.98 %
Cyclist (3D Detection) 69.10 % 53.53 % 47.48 %
Cyclist (Bird's Eye View) 72.93 % 57.77 % 51.03 %
This table as LaTeX


2D object detection results.
This figure as: png eps txt gnuplot



Orientation estimation results.
This figure as: png eps txt gnuplot



3D object detection results.
This figure as: png eps txt gnuplot



Bird's eye view results.
This figure as: png eps txt gnuplot



2D object detection results.
This figure as: png eps txt gnuplot



Orientation estimation results.
This figure as: png eps txt gnuplot



3D object detection results.
This figure as: png eps txt gnuplot



Bird's eye view results.
This figure as: png eps txt gnuplot



2D object detection results.
This figure as: png eps txt gnuplot



Orientation estimation results.
This figure as: png eps txt gnuplot



3D object detection results.
This figure as: png eps txt gnuplot



Bird's eye view results.
This figure as: png eps txt gnuplot




eXTReMe Tracker