SADA: Semantic Adversarial Diagnostic Attacks for Autonomous Applications
KAUST DepartmentElectrical Engineering Program
Computer, Electrical and Mathematical Sciences and Engineering (CEMSE) Division
KAUST Grant NumberRGC/3/3570-01-01
Permanent link to this recordhttp://hdl.handle.net/10754/664563
MetadataShow full item record
AbstractOne major factor impeding more widespread adoption of deep neural networks (DNNs) is their lack of robustness, which is essential for safety-critical applications such as autonomous driving. This has motivated much recent work on adversarial attacks for DNNs, which mostly focus on pixel-level perturbations void of semantic meaning. In contrast, we present a general framework for adversarial attacks on trained agents, which covers semantic perturbations to the environment of the agent performing the task as well as pixel-level attacks. To do this, we re-frame the adversarial attack problem as learning a distribution of parameters that always fools the agent. In the semantic case, our proposed adversary (denoted as BBGAN) is trained to sample parameters that describe the environment with which the black-box agent interacts, such that the agent performs its dedicated task poorly in this environment. We apply BBGAN on three different tasks, primarily targeting aspects of autonomous navigation: object detection, self-driving, and autonomous UAV racing. On these tasks, BBGAN can generate failure cases that consistently fool a trained agent.
CitationHamdi, A., Mueller, M., & Ghanem, B. (2020). SADA: Semantic Adversarial Diagnostic Attacks for Autonomous Applications. Proceedings of the AAAI Conference on Artificial Intelligence, 34(07), 10901–10908. doi:10.1609/aaai.v34i07.6722
SponsorsThis work was supported by the King Abdullah University of Science and Technology (KAUST) Office of Sponsored Research under Award No. RGC/3/3570-01-01.
Conference/Event nameAAAI Conference on Artificial Intelligence