Multi-Object Tracking and Segmentation with a Space-Time Memory Network

Conference on Robots and Vision (CRV 2023)
@inproceedings{miah2023multi,
title={Multi-object tracking and segmentation with a space-time memory network},
author={Miah, Mehdi and Bilodeau, Guillaume-Alexandre and Saunier, Nicolas},
booktitle={2023 20th Conference on Robots and Vision (CRV)},
pages={184--193},
year={2023},
organization={IEEE}
}

Mehdi Miah, Guillaume-Alexandre Bilodeau, Nicolas Saunier

Abstract We propose a method for multi-object tracking and segmentation that does not require finetuning or per benchmark hyper-parameter selection. The proposed tracker, MeNToS, addresses particularly the data association problem. Indeed, the recently introduced HOTA metric, which has a better alignment than the MOTA with the human visual assessment by evenly balancing detections and associations quality, has shown that improvements are still needed for data association. After creating tracklets using instance segmentation and optical flow, the proposed method relies on a space-time memory network originally developed for one-shot video object segmentation to improve the association of tracklets with temporal gaps. We evaluate our tracker on KITTIMOTS and MOTSChallenge and show the benefit of our data association strategy with the HOTA metric. Additional ablation studies demonstrate that our approach using a space-time memory network gives better and more robust long-term association than those based on a re-identification network.
Conference paper
Comments
BibTeX
GitHub
Poster
Video
arXiv
Acknowledgements
We acknowledge the support of the Natural Sciences and Engineering Research Council of Canada (NSERC), [DGDND-2020-04633 and DG individual 06115-2017].