Method

XT-PartA2 [XT-PartA2]


Submitted on 28 Feb. 2024 02:09 by
Xiao Tian (USTC)

Running time:0.1 s
Environment:GPU @ >3.5 Ghz (Python)

Method Description:
Add attention layer to PartA2, train based on MMDetection3D V1.4.0.
Parameters:
Same with PartA2 in MMDetection3D.
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) 95.38 % 91.92 % 89.23 %
Car (Orientation) 95.21 % 91.56 % 88.74 %
Car (3D Detection) 85.54 % 75.56 % 71.02 %
Car (Bird's Eye View) 90.89 % 87.08 % 82.70 %
Pedestrian (Detection) 61.04 % 51.93 % 49.32 %
Pedestrian (Orientation) 57.11 % 47.80 % 45.03 %
Pedestrian (3D Detection) 46.24 % 38.22 % 35.51 %
Pedestrian (Bird's Eye View) 50.62 % 42.68 % 40.25 %
Cyclist (Detection) 87.40 % 76.70 % 71.40 %
Cyclist (Orientation) 87.18 % 76.18 % 70.88 %
Cyclist (3D Detection) 76.88 % 62.82 % 56.99 %
Cyclist (Bird's Eye View) 81.41 % 67.40 % 61.92 %
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