Simple Online and Realtime Tracking with a Deep Association Metric

TL;DR

This paper enhances the SORT tracking algorithm by integrating a deep appearance metric, significantly reducing identity switches and maintaining high processing speeds for real-time multi-object tracking.

Contribution

It introduces a deep learned association metric into SORT, enabling more robust tracking through occlusions and appearance changes, with efficient offline training and online association.

Findings

Reduced identity switches by 45%

Achieved high frame rate performance

Improved long-term object tracking

Abstract

Simple Online and Realtime Tracking (SORT) is a pragmatic approach to multiple object tracking with a focus on simple, effective algorithms. In this paper, we integrate appearance information to improve the performance of SORT. Due to this extension we are able to track objects through longer periods of occlusions, effectively reducing the number of identity switches. In spirit of the original framework we place much of the computational complexity into an offline pre-training stage where we learn a deep association metric on a large-scale person re-identification dataset. During online application, we establish measurement-to-track associations using nearest neighbor queries in visual appearance space. Experimental evaluation shows that our extensions reduce the number of identity switches by 45%, achieving overall competitive performance at high frame rates.