R-PGA: Robust Physical Adversarial Camouflage Generation via Relightable 3D Gaussian Splatting

1. 1Identify the limitations of current physical adversarial camouflage methods
2. 2Develop a relightable 3D Gaussian splatting approach to generate robust camouflage
3. 3Evaluate the effectiveness of R-PGA in complex dynamic scenarios

Computer vision engineers and researchers on autonomous driving teams can benefit from this research to improve the robustness of their systems against physical adversarial attacks

💡 R-PGA can generate camouflage that generalizes across diverse geometric and radiometric variations

💡 R-PGA: robust physical adversarial camouflage generation via relightable 3D Gaussian splatting