Robust Visual Tracking via Statistical Positive Sample Generation and Gradient Aware Learning

TL;DR

This paper introduces SPGA, a novel visual tracking approach that enhances positive sample diversity and balances gradient contributions, leading to improved robustness and accuracy in challenging scenarios.

Contribution

The paper proposes a statistical positive sample generation method and a gradient aware loss to address sample diversity and gradient imbalance in CNN-based trackers.

Findings

SPGA outperforms several state-of-the-art trackers on benchmark datasets.

The statistical positive sample generation enriches training data diversity.

Gradient aware learning improves model robustness against appearance variations.

Abstract

In recent years, Convolutional Neural Network (CNN) based trackers have achieved state-of-the-art performance on multiple benchmark datasets. Most of these trackers train a binary classifier to distinguish the target from its background. However, they suffer from two limitations. Firstly, these trackers cannot effectively handle significant appearance variations due to the limited number of positive samples. Secondly, there exists a significant imbalance of gradient contributions between easy and hard samples, where the easy samples usually dominate the computation of gradient. In this paper, we propose a robust tracking method via Statistical Positive sample generation and Gradient Aware learning (SPGA) to address the above two limitations. To enrich the diversity of positive samples, we present an effective and efficient statistical positive sample generation algorithm to generate positive samples in the feature space. Furthermore, to handle the issue of imbalance between easy and hard samples, we propose a gradient sensitive loss to harmonize the gradient contributions between easy and hard samples. Extensive experiments on three challenging benchmark datasets including OTB50, OTB100 and VOT2016 demonstrate that the proposed SPGA performs favorably against several state-of-the-art trackers.