Adversarial Reinforced Instruction Attacker for Robust Vision-Language Navigation

TL;DR

This paper introduces DR-Attacker, an adversarial training method that enhances the robustness of vision-language navigation models by dynamically attacking and improving instruction understanding during navigation tasks.

Contribution

It proposes a novel adversarial attacking framework using reinforcement learning to generate hard instruction samples, improving navigator robustness in vision-language tasks.

Findings

Outperforms state-of-the-art methods on VLN and NDH tasks.

Effectively attacks crucial instruction information at different timesteps.

Enhances navigation robustness through adversarial training.

Abstract

Language instruction plays an essential role in the natural language grounded navigation tasks. However, navigators trained with limited human-annotated instructions may have difficulties in accurately capturing key information from the complicated instruction at different timesteps, leading to poor navigation performance. In this paper, we exploit to train a more robust navigator which is capable of dynamically extracting crucial factors from the long instruction, by using an adversarial attacking paradigm. Specifically, we propose a Dynamic Reinforced Instruction Attacker (DR-Attacker), which learns to mislead the navigator to move to the wrong target by destroying the most instructive information in instructions at different timesteps. By formulating the perturbation generation as a Markov Decision Process, DR-Attacker is optimized by the reinforcement learning algorithm to generate perturbed instructions sequentially during the navigation, according to a learnable attack score. Then, the perturbed instructions, which serve as hard samples, are used for improving the robustness of the navigator with an effective adversarial training strategy and an auxiliary self-supervised reasoning task. Experimental results on both Vision-and-Language Navigation (VLN) and Navigation from Dialog History (NDH) tasks show the superiority of our proposed method over state-of-the-art methods. Moreover, the visualization analysis shows the effectiveness of the proposed DR-Attacker, which can successfully attack crucial information in the instructions at different timesteps. Code is available at https://github.com/expectorlin/DR-Attacker.