Multi-object tracking by mutual supervision of CNN and particle filter. Xia, Y., Qu, S., Goudos, S., Bai, Y., & Wan, S. Personal and Ubiquitous Computing, 2019.
doi  abstract   bibtex   
© 2019, Springer-Verlag London Ltd., part of Springer Nature. In the multi-object tracking process, a long-term tracking algorithm for traffic scene based on deep learning is proposed to handle several challenging problems, such as the complex variation of background illumination, change of pixel due to partial occlusion, cumulative error and the short-term disappearance of the target. Firstly, we train a CNN to identify and determine the target bounding box in a traffic scene. Secondly, we use a particle filter (PF) as the tracker to implement the preliminary multi-object tracking. Finally, the multi-object tracking trajectory then is generated by the mutual supervision of the PF tracker and CNN detector. In order to evaluate the experimental results, we use the forward-backward (FB) error of our tracker at a certain moment. The experimental results show that the method can track single and multi-objects in long-term tracking in real-time. For the situation of target disappearance and reappearance, the proposed algorithm can also recover its long-term tracking.
@article{
 title = {Multi-object tracking by mutual supervision of CNN and particle filter},
 type = {article},
 year = {2019},
 keywords = {Intelligent transportation systems,Multi-object tracking,Neural networks,Object detection,Particle filters},
 id = {5ad7bec1-1808-3700-9389-9d3919e40da2},
 created = {2020-02-29T16:57:43.783Z},
 file_attached = {false},
 profile_id = {c69aa657-d754-373c-91b7-64154b7d5d91},
 last_modified = {2022-05-12T20:32:46.308Z},
 read = {false},
 starred = {false},
 authored = {true},
 confirmed = {false},
 hidden = {false},
 citation_key = {Xia2019},
 private_publication = {false},
 abstract = {© 2019, Springer-Verlag London Ltd., part of Springer Nature. In the multi-object tracking process, a long-term tracking algorithm for traffic scene based on deep learning is proposed to handle several challenging problems, such as the complex variation of background illumination, change of pixel due to partial occlusion, cumulative error and the short-term disappearance of the target. Firstly, we train a CNN to identify and determine the target bounding box in a traffic scene. Secondly, we use a particle filter (PF) as the tracker to implement the preliminary multi-object tracking. Finally, the multi-object tracking trajectory then is generated by the mutual supervision of the PF tracker and CNN detector. In order to evaluate the experimental results, we use the forward-backward (FB) error of our tracker at a certain moment. The experimental results show that the method can track single and multi-objects in long-term tracking in real-time. For the situation of target disappearance and reappearance, the proposed algorithm can also recover its long-term tracking.},
 bibtype = {article},
 author = {Xia, Y. and Qu, S. and Goudos, S. and Bai, Y. and Wan, S.},
 doi = {10.1007/s00779-019-01278-1},
 journal = {Personal and Ubiquitous Computing}
}

Downloads: 0