Enhanced pedestrian detection using deep learning based semantic image segmentation
Author(s)
Liu, T
Stathaki, T
Type
Conference Paper
Abstract
Pedestrian detection and semantic segmentation are
highly correlated tasks which can be jointly used for better
performance. In this paper, we propose a pedestrian detection
method making use of semantic labeling to improve pedestrian
detection results. A deep learning based semantic segmentation
method is used to pixel-wise label images into 11 common classes.
Semantic segmentation results which encodes high-level image
representation are used as additional feature channels to be
integrated with the low-level HOG+LUV features. Some false
positives, such as falsely detected pedestrians located on a tree,
can be easier eliminated by making use of the semantic cues.
Boosted forest is used for training the integrated feature channels
in a cascaded manner for hard negatives mining. Experiments
on the Caltech-USA pedestrian dataset show improvements on
detection accuracy by using the additional semantic cues.
highly correlated tasks which can be jointly used for better
performance. In this paper, we propose a pedestrian detection
method making use of semantic labeling to improve pedestrian
detection results. A deep learning based semantic segmentation
method is used to pixel-wise label images into 11 common classes.
Semantic segmentation results which encodes high-level image
representation are used as additional feature channels to be
integrated with the low-level HOG+LUV features. Some false
positives, such as falsely detected pedestrians located on a tree,
can be easier eliminated by making use of the semantic cues.
Boosted forest is used for training the integrated feature channels
in a cascaded manner for hard negatives mining. Experiments
on the Caltech-USA pedestrian dataset show improvements on
detection accuracy by using the additional semantic cues.
Date Issued
2017-11-07
Date Acceptance
2017-06-28
Citation
IEEE, 2017
Publisher
IEEE
Journal / Book Title
IEEE
Copyright Statement
© 2017 IEEE. Personal use of this material is permitted. Permission from IEEE must be obtained for all other uses, in any current or future media, including reprinting/republishing this material for advertising or promotional purposes, creating new collective works, for resale or redistribution to servers or lists, or reuse of any copyrighted component of this work in other works.
Sponsor
Commission of the European Communities
Research and Innovation Staff Exchange (RISE)
Identifier
https://ieeexplore.ieee.org/document/8096045
Grant Number
691218
H2020-MSCA-RISE-2015
Source
Digital Signal Processing (DSP) 2017
Subjects
Science & Technology
Technology
Computer Science, Interdisciplinary Applications
Engineering, Electrical & Electronic
Telecommunications
Computer Science
Engineering
Pedestrian detection
semantic segmentation
pixel-wise image labeling
filtered feature channels
Publication Status
Accepted
Start Date
2017-08-21
Finish Date
2017-08-25
Coverage Spatial
London, UK
Date Publish Online
2017-11-07