UHP-SOT: An Unsupervised High-Performance Single Object Tracker

TL;DR

UHP-SOT is an unsupervised single object tracking method that combines appearance modeling, background motion, and trajectory prediction to outperform previous methods and rival supervised deep learning trackers in accuracy and speed.

Contribution

It introduces two novel modules for background motion modeling and trajectory-based prediction to enhance unsupervised tracking performance.

Findings

Outperforms all previous unsupervised SOT methods

Achieves comparable performance to supervised deep-learning trackers

Operates at 22.7-32.0 FPS on CPU

Abstract

An unsupervised online object tracking method that exploits both foreground and background correlations is proposed and named UHP-SOT (Unsupervised High-Performance Single Object Tracker) in this work. UHP-SOT consists of three modules: 1) appearance model update, 2) background motion modeling, and 3) trajectory-based box prediction. A state-of-the-art discriminative correlation filters (DCF) based tracker is adopted by UHP-SOT as the first module. We point out shortcomings of using the first module alone such as failure in recovering from tracking loss and inflexibility in object box adaptation and then propose the second and third modules to overcome them. Both are novel in single object tracking (SOT). We test UHP-SOT on two popular object tracking benchmarks, TB-50 and TB-100, and show that it outperforms all previous unsupervised SOT methods, achieves a performance comparable with the best supervised deep-learning-based SOT methods, and operates at a fast speed (i.e. 22.7-32.0 FPS on a CPU).