A Multilevel Strategy to Improve People Tracking in a Real-World Scenario

TL;DR

This paper introduces a multilevel tracking strategy that enhances people tracking accuracy in a real-world scenario by combining state-of-the-art trackers, validated on a new large-scale dataset from surveillance footage.

Contribution

The paper proposes a novel multilevel hierarchy approach that improves ID association in people tracking, demonstrated on a newly created real-world dataset with significant accuracy gains.

Findings

IDF1 score increased by up to 9.5%

Improved tracking accuracy across multiple metrics

Validated on a large-scale real-world dataset

Abstract

The Pal\'acio do Planalto, office of the President of Brazil, was invaded by protesters on January 8, 2023. Surveillance videos taken from inside the building were subsequently released by the Brazilian Supreme Court for public scrutiny. We used segments of such footage to create the UFPR-Planalto801 dataset for people tracking and re-identification in a real-world scenario. This dataset consists of more than 500,000 images. This paper presents a tracking approach targeting this dataset. The method proposed in this paper relies on the use of known state-of-the-art trackers combined in a multilevel hierarchy to correct the ID association over the trajectories. We evaluated our method using IDF1, MOTA, MOTP and HOTA metrics. The results show improvements for every tracker used in the experiments, with IDF1 score increasing by a margin up to 9.5%.