Robust Deep Reinforcement Learning in Robotics via Adaptive Gradient-Masked Adversarial Attacks
Fuente:
arXiv
Enregistré dans:
| Auteurs principaux: | , , , , , , , , , , |
|---|---|
| Format: | Preprint |
| Publié: |
2025
|
| Sujets: | |
| Accès en ligne: | |
| Tags: |
Ajouter un tag
Pas de tags, Soyez le premier à ajouter un tag!
|
| _version_ | 1866908285770137600 |
|---|---|
| author | Zhang, Zongyuan Duan, Tianyang Lin, Zheng Huang, Dong Fang, Zihan Sun, Zekai Xiong, Ling Liang, Hongbin Cui, Heming Cui, Yong Gao, Yue |
| author_facet | Zhang, Zongyuan Duan, Tianyang Lin, Zheng Huang, Dong Fang, Zihan Sun, Zekai Xiong, Ling Liang, Hongbin Cui, Heming Cui, Yong Gao, Yue |
| contents | Deep reinforcement learning (DRL) has emerged as a promising approach for robotic control, but its realworld deployment remains challenging due to its vulnerability to environmental perturbations. Existing white-box adversarial attack methods, adapted from supervised learning, fail to effectively target DRL agents as they overlook temporal dynamics and indiscriminately perturb all state dimensions, limiting their impact on long-term rewards. To address these challenges, we propose the Adaptive Gradient-Masked Reinforcement (AGMR) Attack, a white-box attack method that combines DRL with a gradient-based soft masking mechanism to dynamically identify critical state dimensions and optimize adversarial policies. AGMR selectively allocates perturbations to the most impactful state features and incorporates a dynamic adjustment mechanism to balance exploration and exploitation during training. Extensive experiments demonstrate that AGMR outperforms state-of-the-art adversarial attack methods in degrading the performance of the victim agent and enhances the victim agent's robustness through adversarial defense mechanisms. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2503_20844 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | Robust Deep Reinforcement Learning in Robotics via Adaptive Gradient-Masked Adversarial Attacks Zhang, Zongyuan Duan, Tianyang Lin, Zheng Huang, Dong Fang, Zihan Sun, Zekai Xiong, Ling Liang, Hongbin Cui, Heming Cui, Yong Gao, Yue Machine Learning Artificial Intelligence Networking and Internet Architecture Robotics Deep reinforcement learning (DRL) has emerged as a promising approach for robotic control, but its realworld deployment remains challenging due to its vulnerability to environmental perturbations. Existing white-box adversarial attack methods, adapted from supervised learning, fail to effectively target DRL agents as they overlook temporal dynamics and indiscriminately perturb all state dimensions, limiting their impact on long-term rewards. To address these challenges, we propose the Adaptive Gradient-Masked Reinforcement (AGMR) Attack, a white-box attack method that combines DRL with a gradient-based soft masking mechanism to dynamically identify critical state dimensions and optimize adversarial policies. AGMR selectively allocates perturbations to the most impactful state features and incorporates a dynamic adjustment mechanism to balance exploration and exploitation during training. Extensive experiments demonstrate that AGMR outperforms state-of-the-art adversarial attack methods in degrading the performance of the victim agent and enhances the victim agent's robustness through adversarial defense mechanisms. |
| title | Robust Deep Reinforcement Learning in Robotics via Adaptive Gradient-Masked Adversarial Attacks |
| topic | Machine Learning Artificial Intelligence Networking and Internet Architecture Robotics |
| url | https://arxiv.org/abs/2503.20844 |