Physics-inspired deepfake detector uses Hamiltonian dynamics for stability analysis

Researchers have developed a novel method for detecting deepfakes by analyzing their underlying stability rather than just visual patterns. This approach, called Hamiltonian Action Anomaly Detection (HAAD), models images on a potential energy surface, hypothesizing that real images reside in stable, low-energy states while deepfakes occupy unstable, high-energy states. By simulating Hamiltonian dynamics, HAAD quantifies these differences through trajectory statistics, outperforming existing methods on cross-dataset benchmarks. AI

Summary written by gemini-2.5-flash-lite from 1 source. How we write summaries →

IMPACT This physics-inspired approach to deepfake detection could offer a more robust defense against evolving generative AI.

RANK_REASON This is a research paper detailing a new method for deepfake detection. [lever_c_demoted from research: ic=1 ai=1.0]

Read on arXiv cs.CV →

• arXiv
• Hamiltonian Dynamics
• Hamiltonian Action Anomaly Detection
• HAAD

• paper
• other