AI & ML Practical Magic

A cheap plastic sheet on your camera lens creates a "fingerprint" that even the smartest AI can't fake.

April 3, 2026

Original Paper

Moiré Video Authentication: A Physical Signature Against AI Video Generation

Yuan Qing, Kunyu Zheng, Lingxiao Li, Boqing Gong, Chang Xiao

arXiv · 2604.01654

The Takeaway

By placing a cheap optical pattern in front of a camera, researchers created a physical signature that follows strict laws of light. AI video generators can mimic how things look, but they fail to recreate the exact mathematical physics of these interference patterns, making real video instantly verifiable.

From the abstract

Recent advances in video generation have made AI-synthesized content increasingly difficult to distinguish from real footage. We propose a physics-based authentication signature that real cameras produce naturally, but that generative models cannot faithfully reproduce. Our approach exploits the Moiré effect: the interference fringes formed when a camera views a compact two-layer grating structure. We derive the Moiré motion invariant, showing that fringe phase and grating image displacement are