LRR: Language-Driven Resamplable Continuous Representation against Adversarial Tracking Attacks

TL;DR

This paper introduces a language-driven continuous representation for visual object tracking that enhances robustness against adversarial attacks while maintaining high accuracy on clean data.

Contribution

It proposes a novel spatial-temporal continuous representation guided by semantic text to defend against adversarial tracking attacks.

Findings

Achieves around 90% relative improvement under adversarial attacks on UAV123

Maintains high accuracy on clean data

Successfully defends against multiple SOTA adversarial attacks

Abstract

Visual object tracking plays a critical role in visual-based autonomous systems, as it aims to estimate the position and size of the object of interest within a live video. Despite significant progress made in this field, state-of-the-art (SOTA) trackers often fail when faced with adversarial perturbations in the incoming frames. This can lead to significant robustness and security issues when these trackers are deployed in the real world. To achieve high accuracy on both clean and adversarial data, we propose building a spatial-temporal continuous representation using the semantic text guidance of the object of interest. This novel continuous representation enables us to reconstruct incoming frames to maintain semantic and appearance consistency with the object of interest and its clean counterparts. As a result, our proposed method successfully defends against different SOTA adversarial tracking attacks while maintaining high accuracy on clean data. In particular, our method significantly increases tracking accuracy under adversarial attacks with around 90% relative improvement on UAV123, which is even higher than the accuracy on clean data.