Learning Reinforced Attentional Representation for End-to-End Visual Tracking

TL;DR

This paper introduces an end-to-end neural network with hierarchical attention and online correlation filter updates for improved visual object tracking, achieving high accuracy and efficiency.

Contribution

It proposes a novel hierarchical attentional module with LSTM and MLPs, and integrates a contextual attentional correlation filter for end-to-end training and online adaptation.

Findings

Effective in discriminating and localizing targets

Achieves competitive accuracy on benchmark datasets

Operates with high computational efficiency

Abstract

Although numerous recent tracking approaches have made tremendous advances in the last decade, achieving high-performance visual tracking remains a challenge. In this paper, we propose an end-to-end network model to learn reinforced attentional representation for accurate target object discrimination and localization. We utilize a novel hierarchical attentional module with long short-term memory and multi-layer perceptrons to leverage both inter- and intra-frame attention to effectively facilitate visual pattern emphasis. Moreover, we incorporate a contextual attentional correlation filter into the backbone network to make our model trainable in an end-to-end fashion. Our proposed approach not only takes full advantage of informative geometries and semantics but also updates correlation filters online without fine-tuning the backbone network to enable the adaptation of variations in the target object's appearance. Extensive experiments conducted on several popular benchmark datasets demonstrate that our proposed approach is effective and computationally efficient.