Method

CSHOT on Laser Reflectance [la] [CSoR]


Submitted on 24 Jul. 2017 13:24 by
David Stutz (Max Planck Institute for Intelligent Systems)

Running time:3.5 s
Environment:4 cores @ >3.5 Ghz (Python + C/C++)

Method Description:
Estimation of object position with k-Means, AdaBoost
and kNN based on CSHOT features computed for LIDAR
data including laser reflectance.

Originally submitted by Leonard Plotkin (KIT).
Parameters:
radius=2.0m
Latex Bibtex:
@MastersThesis{Plotkin2015,
author = {Leonard Plotkin},
title = {PyDriver: Entwicklung eines Frameworks
für räumliche Detektion und Klassifikation von
Objekten in Fahrzeugumgebung},
school = {Karlsruhe Institute of Technology},
type = {Bachelor's Thesis ({Studienarbeit})},
address = {Germany},
month = MAR,
year = 2015,
url =
{https://github.com/lpltk/pydriver/releases/downloa
d/v0.1.0/PyDriver_thesis.pdf}
}

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 (3D Detection) 5.61 % 4.06 % 3.17 %
Car (Bird's Eye View) 18.67 % 13.07 % 10.34 %
This table as LaTeX


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