Online multi-object tracking via labeled random finite set with appearance learning

© 2017 IEEE. In this paper, a novel approach to online multi-object tracking is proposed via Labeled Random Finite Sets (RFS) combined with appearance learning. The Labeled RFS formulation of the multi-object state naturally accommodates a time-varying number of objects, track labels, and false posi...

Full description

Bibliographic Details
Main Author: Kim, Du Yong
Format: Conference Paper
Published: 2017
Online Access:http://hdl.handle.net/20.500.11937/66632
_version_ 1848761357019643904
author Kim, Du Yong
author_facet Kim, Du Yong
author_sort Kim, Du Yong
building Curtin Institutional Repository
collection Online Access
description © 2017 IEEE. In this paper, a novel approach to online multi-object tracking is proposed via Labeled Random Finite Sets (RFS) combined with appearance learning. The Labeled RFS formulation of the multi-object state naturally accommodates a time-varying number of objects, track labels, and false positive rejection in a single Bayesian framework. The proposed algorithm exploits appearance feature information for the purpose of learning an object's appearance model, and uses this additional information in the construction an augmented likelihood which improves performance and facilitates track re-initialization. This approach enhances the baseline tracking algorithm and shows better performance with respect to mis-detections, occlusions and false track rejection. Competitive tracking results are shown compared to state-of-the-art algorithms on PETS benchmark [1] video datasets.
first_indexed 2025-11-14T10:30:23Z
format Conference Paper
id curtin-20.500.11937-66632
institution Curtin University Malaysia
institution_category Local University
last_indexed 2025-11-14T10:30:23Z
publishDate 2017
recordtype eprints
repository_type Digital Repository
spelling curtin-20.500.11937-666322018-05-14T06:17:52Z Online multi-object tracking via labeled random finite set with appearance learning Kim, Du Yong © 2017 IEEE. In this paper, a novel approach to online multi-object tracking is proposed via Labeled Random Finite Sets (RFS) combined with appearance learning. The Labeled RFS formulation of the multi-object state naturally accommodates a time-varying number of objects, track labels, and false positive rejection in a single Bayesian framework. The proposed algorithm exploits appearance feature information for the purpose of learning an object's appearance model, and uses this additional information in the construction an augmented likelihood which improves performance and facilitates track re-initialization. This approach enhances the baseline tracking algorithm and shows better performance with respect to mis-detections, occlusions and false track rejection. Competitive tracking results are shown compared to state-of-the-art algorithms on PETS benchmark [1] video datasets. 2017 Conference Paper http://hdl.handle.net/20.500.11937/66632 10.1109/ICCAIS.2017.8217572 restricted
spellingShingle Kim, Du Yong
Online multi-object tracking via labeled random finite set with appearance learning
title Online multi-object tracking via labeled random finite set with appearance learning
title_full Online multi-object tracking via labeled random finite set with appearance learning
title_fullStr Online multi-object tracking via labeled random finite set with appearance learning
title_full_unstemmed Online multi-object tracking via labeled random finite set with appearance learning
title_short Online multi-object tracking via labeled random finite set with appearance learning
title_sort online multi-object tracking via labeled random finite set with appearance learning
url http://hdl.handle.net/20.500.11937/66632