Large Margin Object Tracking with Circulant Feature Maps

TL;DR

This paper introduces a fast, large margin object tracking method combining structured output SVM advantages with correlation filters, incorporating multimodal detection and feedback mechanisms for improved accuracy and real-time performance.

Contribution

It presents a novel large margin tracking algorithm that integrates correlation filters and multimodal detection, achieving high accuracy and speed with robustness against model drift.

Findings

Achieves over 80 fps in tracking speed.

Outperforms state-of-the-art algorithms on benchmark sequences.

Demonstrates robustness with both handcrafted and deep CNN features.

Abstract

Structured output support vector machine (SVM) based tracking algorithms have shown favorable performance recently. Nonetheless, the time-consuming candidate sampling and complex optimization limit their real-time applications. In this paper, we propose a novel large margin object tracking method which absorbs the strong discriminative ability from structured output SVM and speeds up by the correlation filter algorithm significantly. Secondly, a multimodal target detection technique is proposed to improve the target localization precision and prevent model drift introduced by similar objects or background noise. Thirdly, we exploit the feedback from high-confidence tracking results to avoid the model corruption problem. We implement two versions of the proposed tracker with the representations from both conventional hand-crafted and deep convolution neural networks (CNNs) based features to validate the strong compatibility of the algorithm. The experimental results demonstrate that the proposed tracker performs superiorly against several state-of-the-art algorithms on the challenging benchmark sequences while runs at speed in excess of 80 frames per second. The source code and experimental results will be made publicly available.