MOT20: A benchmark for multi object tracking in crowded scenes

TL;DR

The paper introduces MOT20, a new benchmark dataset with crowded scenes for evaluating multi-object tracking methods, addressing the challenge of tracking in extremely crowded environments.

Contribution

It provides a new dataset with 8 challenging crowded sequences to evaluate multi-object tracking algorithms in dense scenarios.

Findings

Benchmark facilitates evaluation of tracking methods in crowded scenes

Highlights the difficulty of tracking in extremely crowded environments

Supports development of more robust multi-object tracking algorithms

Abstract

Standardized benchmarks are crucial for the majority of computer vision applications. Although leaderboards and ranking tables should not be over-claimed, benchmarks often provide the most objective measure of performance and are therefore important guides for research. The benchmark for Multiple Object Tracking, MOTChallenge, was launched with the goal to establish a standardized evaluation of multiple object tracking methods. The challenge focuses on multiple people tracking, since pedestrians are well studied in the tracking community, and precise tracking and detection has high practical relevance. Since the first release, MOT15, MOT16, and MOT17 have tremendously contributed to the community by introducing a clean dataset and precise framework to benchmark multi-object trackers. In this paper, we present our MOT20benchmark, consisting of 8 new sequences depicting very crowded challenging scenes. The benchmark was presented first at the 4thBMTT MOT Challenge Workshop at the Computer Vision and Pattern Recognition Conference (CVPR) 2019, and gives to chance to evaluate state-of-the-art methods for multiple object tracking when handling extremely crowded scenarios.