Saved in:
| Main Authors: | , , , , , , , , |
|---|---|
| Format: | Preprint |
| Published: |
2022
|
| Subjects: | |
| Online Access: | https://arxiv.org/abs/2212.06123 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866912603707539456 |
|---|---|
| author | Demontis, Ambra Gupta, Srishti Pintor, Maura Demetrio, Luca Grosse, Kathrin Lin, Hsiao-Ying Fang, Chengfang Biggio, Battista Roli, Fabio |
| author_facet | Demontis, Ambra Gupta, Srishti Pintor, Maura Demetrio, Luca Grosse, Kathrin Lin, Hsiao-Ying Fang, Chengfang Biggio, Battista Roli, Fabio |
| contents | Reinforcement learning (RL) enables agents to learn optimal behaviors through interaction with their environment and has been increasingly deployed in safety-critical applications, including autonomous driving. Despite its promise, RL is susceptible to attacks designed either to compromise policy learning or to induce erroneous decisions by trained agents. Although the literature on RL security has grown rapidly and several surveys exist, existing categorizations often fall short in guiding the selection of appropriate defenses for specific systems. In this work, we present a comprehensive survey of 86 recent studies on RL security, addressing these limitations by systematically categorizing attacks and defenses according to defined threat models and single- versus multi-agent settings. Furthermore, we examine the relevance and applicability of state-of-the-art attacks and defense mechanisms within the context of autonomous driving, providing insights to inform the design of robust RL systems. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2212_06123 |
| institution | arXiv |
| publishDate | 2022 |
| record_format | arxiv |
| spellingShingle | Security of Deep Reinforcement Learning for Autonomous Driving: A Survey Demontis, Ambra Gupta, Srishti Pintor, Maura Demetrio, Luca Grosse, Kathrin Lin, Hsiao-Ying Fang, Chengfang Biggio, Battista Roli, Fabio Machine Learning Robotics Reinforcement learning (RL) enables agents to learn optimal behaviors through interaction with their environment and has been increasingly deployed in safety-critical applications, including autonomous driving. Despite its promise, RL is susceptible to attacks designed either to compromise policy learning or to induce erroneous decisions by trained agents. Although the literature on RL security has grown rapidly and several surveys exist, existing categorizations often fall short in guiding the selection of appropriate defenses for specific systems. In this work, we present a comprehensive survey of 86 recent studies on RL security, addressing these limitations by systematically categorizing attacks and defenses according to defined threat models and single- versus multi-agent settings. Furthermore, we examine the relevance and applicability of state-of-the-art attacks and defense mechanisms within the context of autonomous driving, providing insights to inform the design of robust RL systems. |
| title | Security of Deep Reinforcement Learning for Autonomous Driving: A Survey |
| topic | Machine Learning Robotics |
| url | https://arxiv.org/abs/2212.06123 |