FlagThis

Abstract

Deep neural networks have demonstrated excellent performance in SAR object detection tasks, but remain susceptible to adversarial attacks. Existing SAR-specific attack methods can effectively deceive detectors while introducing noticeable perturbations to the digital domain, neglecting physical implementation constraints for attacking SAR object detection models. In this paper, a novel adversarial attack named Adversarial Attenuation Patch (AAP) is proposed, which employs an energy-constrained optimization strategy coupled with an attenuation-based deployment framework to achieve a seamless balance between attack effectiveness and stealthiness. AAP exhibits strong potential for physical realization by aligning with signal-level electronic jamming. Experimental results show that AAP effectively degrades detection performance while preserving high imperceptibility and shows favorable transferability across different models. This study provides a new perspective for physically realizable adversarial attacks against SAR object detection. The source code is made available at https://github.com/boremycin/SAAP.