Method

Enhanced-VoxelNet [E-VoxelNet]
[Anonymous Submission]

Submitted on 4 Aug. 2019 13:09 by
[Anonymous Submission]

Running time:0.1 s
Environment:GPU @ 2.5 Ghz (Python)

Method Description:
It is implemented based on VoxelNet with DropBlock
and Sparse Convolution enhancement.
Parameters:
\gamma=2.0
Latex Bibtex:

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) 70.24 % 63.00 % 55.94 %
Car (Orientation) 30.85 % 28.03 % 25.39 %
Car (3D Detection) 67.73 % 54.33 % 47.70 %
Car (Bird's Eye View) 77.08 % 68.34 % 60.15 %
This table as LaTeX


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



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



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



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




eXTReMe Tracker