Intentional Electromagnetic Interference Attacks on Facial Recognition

arXiv:2607.15512 · cs.CV, cs.CR, cs.LG · Submitted 2026-07-16 · Read on arXiv

Tyler Fitzsimmons, Adam Czajka

cs.CV, cs.CR, cs.LG

Submitted: 2026-07-16

Comments: To be published in IEEE/IAPR IJCB (International Joint Conference on Biometrics) 2026

Code: https://github.com/CVRL/EMI-Attacks-Face

License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/

Terminology

• Loss Function — 10× in this paper · explained in 2 episodes
• Robustness — 5× in this paper · explained in 5 episodes
• Benchmark — 4× in this paper · explained in 2 episodes
• Generalization — 4× in this paper · explained in 13 episodes
• Cosine Similarity — 3× in this paper · explained in 3 episodes
• False Non-Match Rate — 3× in this paper
• Verification — 3× in this paper · explained in 2 episodes
• Accuracy — 2× in this paper · explained in 1 episode
• Distillation — 2× in this paper · explained in 2 episodes
• Embeddings — 2× in this paper · explained in 2 episodes
• Evasion Success Rate — 2× in this paper
• False Match Rate — 2× in this paper
• Multiple Biometric Grand Challenge — 2× in this paper
• Transferability — 2× in this paper · explained in 4 episodes
• Adam — 1× in this paper · explained in 2 episodes
• Adversarial Attacks — 1× in this paper · explained in 1 episode
• Adversarial Examples — 1× in this paper · explained in 2 episodes
• Adversarial Robustness — 1× in this paper · explained in 3 episodes
• Conference on Computer Vision and Pattern Recognition — 1× in this paper
• Explainability — 1× in this paper · explained in 3 episodes
• Fast Fourier Transform — 1× in this paper
• Fine-Tuning — 1× in this paper · explained in 5 episodes
• Focal Plane Array — 1× in this paper
• Foundation Model — 1× in this paper · explained in 3 episodes
• Foundation Models — 1× in this paper · explained in 7 episodes
• Horizontal Timing Control — 1× in this paper
• Optical Image Stabilization — 1× in this paper
• Physical Attack Success — 1× in this paper
• Transformer — 1× in this paper · explained in 2 episodes
• Transformer Architecture — 1× in this paper · explained in 2 episodes

Sources

• YOLOv4: Optimal Speed and Accuracy of Object Detection
• DINOv3
• YOLOv3: An Incremental Improvement
• Rainbow Artifacts from Electromagnetic Signal Injection Attacks on Image Sensors

Related papers

• Loss Knows Best: Detecting Annotation Errors in Videos via Loss Trajectories
• AnchorWeave: World-Consistent Video Generation with Retrieved Local Spatial Memories
• Benchmarking the Robustness of Foundation Models for Mammography under Domain Shift
• MambaX-Net: Dual-Input Mamba-Enhanced Cross-Attention Network for Longitudinal MRI Segmentation
• TeleOCR: Navigating Document Parsing Across Digital and Camera-Captured Documents
• A Survey on Efficient Vision-Language-Action Models