Authors:
Wiebe Van Ranst
1
;
Floris De Smedt
2
and
Toon Goedemé
1
Affiliations:
1
KU Leuven, Belgium
;
2
Robovision BVBA, Belgium
Keyword(s):
Person Detection, ACF, GPU, CUDA, Embedded.
Related
Ontology
Subjects/Areas/Topics:
Applications
;
Applications and Services
;
Computer Vision, Visualization and Computer Graphics
;
Features Extraction
;
Image and Video Analysis
;
Mobile Imaging
;
Pattern Recognition
;
Robotics
;
Shape Representation and Matching
;
Software Engineering
Abstract:
The field of pedestrian detection has come a long way in recent decades. In terms of accuracy, the current
state-of-the-art is hands down reached by Deep Learning methods. However in terms of running speed this is
not always the case, traditional methods are often still faster than their Deep Learning counterparts. This is
especially true on embedded hardware, embedded platforms are often used in applications that require realtime
performance while at same the time having to make do with a limited amount of resources. In this paper
we present a GPU implementation of the ACF pedestrian detector and compare it to current Deep Learning
approaches (YOLO) on both a desktop GPU as well as the Jetson TX2 embedded GPU platform.