OCTrack: Benchmarking the Open-Corpus Multi-Object Tracking

TL;DR

This paper introduces OCTrackB, a comprehensive benchmark for open-corpus multi-object tracking that evaluates models on their ability to localize, associate, and recognize both seen and unseen object categories without category prompts.

Contribution

The paper presents OCTrackB, a large-scale benchmark dataset and a new recognition metric for evaluating open-corpus multi-object tracking methods.

Findings

State-of-the-art methods show varied performance on open-corpus tracking.

OCTrackB provides a less biased, balanced evaluation of base and novel classes.

The new metric improves assessment of generative object recognition.

Abstract

We study a novel yet practical problem of open-corpus multi-object tracking (OCMOT), which extends the MOT into localizing, associating, and recognizing generic-category objects of both seen (base) and unseen (novel) classes, but without the category text list as prompt. To study this problem, the top priority is to build a benchmark. In this work, we build OCTrackB, a large-scale and comprehensive benchmark, to provide a standard evaluation platform for the OCMOT problem. Compared to previous datasets, OCTrackB has more abundant and balanced base/novel classes and the corresponding samples for evaluation with less bias. We also propose a new multi-granularity recognition metric to better evaluate the generative object recognition in OCMOT. By conducting the extensive benchmark evaluation, we report and analyze the results of various state-of-the-art methods, which demonstrate the rationale of OCMOT, as well as the usefulness and advantages of OCTrackB.