Global Instance Tracking: Locating Target More Like Humans

TL;DR

This paper introduces a new global instance tracking task and benchmark to better evaluate and develop trackers that mimic human visual tracking abilities, addressing limitations of existing methods in challenging scenarios.

Contribution

It proposes the GIT task, constructs the VideoCube benchmark, and develops a scientific evaluation procedure based on human capabilities to assess tracking intelligence.

Findings

Significant gap between current trackers and human performance.

Established a challenging benchmark for tracking in complex scenarios.

Provided an online platform with tools and leaderboard for evaluation.

Abstract

Target tracking, the essential ability of the human visual system, has been simulated by computer vision tasks. However, existing trackers perform well in austere experimental environments but fail in challenges like occlusion and fast motion. The massive gap indicates that researches only measure tracking performance rather than intelligence. How to scientifically judge the intelligence level of trackers? Distinct from decision-making problems, lacking three requirements (a challenging task, a fair environment, and a scientific evaluation procedure) makes it strenuous to answer the question. In this article, we first propose the global instance tracking (GIT) task, which is supposed to search an arbitrary user-specified instance in a video without any assumptions about camera or motion consistency, to model the human visual tracking ability. Whereafter, we construct a high-quality and large-scale benchmark VideoCube to create a challenging environment. Finally, we design a scientific evaluation procedure using human capabilities as the baseline to judge tracking intelligence. Additionally, we provide an online platform with toolkit and an updated leaderboard. Although the experimental results indicate a definite gap between trackers and humans, we expect to take a step forward to generate authentic human-like trackers. The database, toolkit, evaluation server, and baseline results are available at http://videocube.aitestunion.com.