Anti-ESIA: Analyzing and Mitigating Impacts of Electromagnetic Signal Injection Attacks

TL;DR

Electromagnetic Signal Injection Attacks (ESIA) pose a hardware-level threat to camera-based perception in intelligent systems. The paper analyzes ESIA via two failure modes—pixel loss and color strips—using a synthetic row-dropping framework and evaluates ViT-B-32 accuracy on $n \in \{0,16,22,46,66,76\}$ with 495 ImageNet validation images, finding that color-strip artifacts account for roughly $0.74$ of the degradation. It formalizes the attack as $Degradation = f(R) + g(R)$ with $R=(r_0,\ldots,r_{n-1})$ and demonstrates that color distortion is more impactful than pixel loss. A lightweight defense based on median interpolation is shown to partially recover performance for moderate attacks but struggles under severe row loss, highlighting the need for more robust reconstruction, color-strip detection, and broader applicability to other domains.

Abstract

Cameras are integral components of many critical intelligent systems. However, a growing threat, known as Electromagnetic Signal Injection Attacks (ESIA), poses a significant risk to these systems, where ESIA enables attackers to remotely manipulate images captured by cameras, potentially leading to malicious actions and catastrophic consequences. Despite the severity of this threat, the underlying reasons for ESIA's effectiveness remain poorly understood, and effective countermeasures are lacking. This paper aims to address these gaps by investigating ESIA from two distinct aspects: pixel loss and color strips. By analyzing these aspects separately on image classification tasks, we gain a deeper understanding of how ESIA can compromise intelligent systems. Additionally, we explore a lightweight solution to mitigate the effects of ESIA while acknowledging its limitations. Our findings provide valuable insights for future research and development in the field of camera security and intelligent systems.

Figures (2)

• Figure 1: ESIA can cause pixel loss and color strips, misleading intelligent systems to make wrong decisions.
• Figure 2: Performance is degraded by ESIA with or without color strips, and performance can be recovered after applying our mitigation method.