Performance Measures and a Data Set for Multi-Target, Multi-Camera Tracking

TL;DR

This paper introduces new performance metrics, a large annotated dataset, and a baseline system for multi-target, multi-camera tracking, addressing current challenges and enabling better evaluation of tracking systems.

Contribution

It presents a novel set of performance measures, the largest annotated dataset to date, and a reference system for multi-camera tracking research.

Findings

Metrics effectively evaluate identity matching performance.

Dataset presents realistic tracking challenges.

Baseline system achieves state-of-the-art performance.

Abstract

To help accelerate progress in multi-target, multi-camera tracking systems, we present (i) a new pair of precision-recall measures of performance that treats errors of all types uniformly and emphasizes correct identification over sources of error; (ii) the largest fully-annotated and calibrated data set to date with more than 2 million frames of 1080p, 60fps video taken by 8 cameras observing more than 2,700 identities over 85 minutes; and (iii) a reference software system as a comparison baseline. We show that (i) our measures properly account for bottom-line identity match performance in the multi-camera setting; (ii) our data set poses realistic challenges to current trackers; and (iii) the performance of our system is comparable to the state of the art.