Method

Transformation-Based 3-D Object Detection via a Spatial Shape Transformer [TSSTDet]


Submitted on 22 Aug. 2023 04:36 by
Hiep Hoang Anh (Soongsil University)

Running time:0.08 s
Environment:1 core @ 2.5 Ghz (C/C++)

Method Description:
Multistage
Parameters:
0.2
Latex Bibtex:
@ARTICLE{10399338,
author={Hoang, Hiep Anh and Bui, Duy Cuong and
Yoo, Myungsik},
journal={IEEE Sensors Journal},
title={TSSTDet: Transformation-Based 3-D Object
Detection via a Spatial Shape Transformer},
year={2024},
volume={24},
number={5},
pages={7126-7139},
keywords={Point cloud compression;Shape;Object
detection;Transformers;Feature extraction;Task
analysis;Detectors;3-D object detection;autonomous
driving;light detection and ranging (LiDAR) point
cloud;vision transformer},
doi={10.1109/JSEN.2024.3350770}}

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) 96.65 % 95.81 % 93.05 %
Car (Orientation) 96.54 % 95.56 % 92.71 %
Car (3D Detection) 91.84 % 85.47 % 80.65 %
Car (Bird's Eye View) 95.80 % 92.11 % 89.23 %
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




eXTReMe Tracker