loading
  • Login
  • Sign-Up

Research.Publish.Connect.

Paper

Authors: Nathanael L. Baisa 1 ; Deepayan Bhowmik 2 and Andrew Wallace 1

Affiliations: 1 Heriot-Watt University, United Kingdom ; 2 Sheffield Hallam University, United Kingdom

ISBN: 978-989-758-227-1

Keyword(s): Visual Tracking, Correlation Filter, CNN Features, Hybrid Features, Online Learning, GM-PHD Filter.

Related Ontology Subjects/Areas/Topics: Computer Vision, Visualization and Computer Graphics ; Motion, Tracking and Stereo Vision ; Tracking and Visual Navigation ; Video Surveillance and Event Detection

Abstract: Tracking a target of interest in crowded environments is a challenging problem, not yet successfully addressed in the literature. In this paper, we propose a new long-term algorithm, learning a discriminative correlation filter and using an online classifier, to track a target of interest in dense video sequences. First, we learn a translational correlation filter using a multi-layer hybrid of convolutional neural networks (CNN) and traditional hand-crafted features. We combine the advantages of both the lower convolutional layer which retains better spatial detail for precise localization, and the higher convolutional layer which encodes semantic information for handling appearance variations. This is integrated with traditional features formed from a histogram of oriented gradients (HOG) and color-naming. Second, we include a re-detection module for overcoming tracking failures due to long-term occlusions by training an incremental (online) SVM on the most confident frames using han d-engineered features. This re-detection module is activated only when the correlation response of the object is below some pre-defined threshold to generate high score detection proposals. Finally, we incorporate a Gaussian mixture probability hypothesis density (GM-PHD) filter to temporally filter high score detection proposals generated from the learned online SVM to find the detection proposal with the maximum weight as the target position estimate by removing the other detection proposals as clutter. Extensive experiments on dense data sets show that our method significantly outperforms state-of-the-art methods. (More)

PDF ImageFull Text

Download
Sign In Guest: Register as new SCITEPRESS user or Join INSTICC now for free.

Sign In SCITEPRESS user: please login.

Sign In INSTICC Members: please login. If not a member yet, Join INSTICC now for free.

PDF ImageMy Papers

You are not signed in, therefore limits apply to your IP address 54.81.210.42. INSTICC members have higher download limits (free membership now)

In the current month:
Recent papers: 1 available of 1 total
2+ years older papers: 2 available of 2 total

Paper citation in several formats:
L. Baisa N., Bhowmik D. and Wallace A. (2017). Long-term Correlation Tracking using Multi-layer Hybrid Features in Dense Environments.In Proceedings of the 12th International Joint Conference on Computer Vision, Imaging and Computer Graphics Theory and Applications - Volume 6: VISAPP, (VISIGRAPP 2017) ISBN 978-989-758-227-1, pages 192-203. DOI: 10.5220/0006117301920203

@conference{visapp17,
author={Nathanael L. Baisa and Deepayan Bhowmik and Andrew Wallace},
title={Long-term Correlation Tracking using Multi-layer Hybrid Features in Dense Environments},
booktitle={Proceedings of the 12th International Joint Conference on Computer Vision, Imaging and Computer Graphics Theory and Applications - Volume 6: VISAPP, (VISIGRAPP 2017)},
year={2017},
pages={192-203},
publisher={ScitePress},
organization={INSTICC},
doi={10.5220/0006117301920203},
isbn={978-989-758-227-1},
}

TY - CONF

JO - Proceedings of the 12th International Joint Conference on Computer Vision, Imaging and Computer Graphics Theory and Applications - Volume 6: VISAPP, (VISIGRAPP 2017)
TI - Long-term Correlation Tracking using Multi-layer Hybrid Features in Dense Environments
SN - 978-989-758-227-1
AU - L. Baisa N.
AU - Bhowmik D.
AU - Wallace A.
PY - 2017
SP - 192
EP - 203
DO - 10.5220/0006117301920203

Sorted by: Show papers

Note: The preferred Subjects/Areas/Topics, listed below for each paper, are those that match the selected paper topics and their ontology superclasses.
More...

Login or register to post comments.

Comments on this Paper: Be the first to review this paper.

Show authors

Note: The preferred Subjects/Areas/Topics, listed below for each author, are those that more frequently used in the author's papers.
More...