PulseAugur
EN
LIVE 10:25:32

New UNDREAM framework enables end-to-end adversarial attacks in simulations

By PulseAugur Editorial · [1 sources] ·

Researchers have developed UNDREAM, a novel software framework designed to bridge the gap between photorealistic simulators and differentiable renderers. This innovation enables end-to-end optimization of adversarial perturbations on 3D objects within simulated environments. UNDREAM offers comprehensive control over environmental factors such as weather, lighting, and camera angles, facilitating the creation of diverse and realistic adversarial attack scenarios. The framework's ability to combine photorealistic simulation with differentiable optimization is expected to advance research in physical adversarial attacks. AI

IMPACT Enables more realistic and comprehensive testing of AI robustness against adversarial attacks in safety-critical applications.

RANK_REASON Academic paper detailing a new software framework for adversarial attacks. [lever_c_demoted from research: ic=1 ai=1.0]

Read on arXiv cs.AI →

AI-generated summary · Google Gemini · from 1 sources. How we write summaries →

New UNDREAM framework enables end-to-end adversarial attacks in simulations

COVERAGE [1]

  1. arXiv cs.AI TIER_1 English(EN) · Mansi Phute, Matthew Hull, Haoran Wang, Alec Helbling, ShengYun Peng, Willian Lunardi, Martin Andreoni, Wenke Lee, Duen Horng Chau ·

    UNDREAM: Bridging Differentiable Rendering and Photorealistic Simulation for End-to-end Adversarial Attacks

    arXiv:2510.16923v3 Announce Type: replace-cross Abstract: Deep learning models deployed in safety critical applications like autonomous driving use simulations to test their robustness against adversarial attacks in realistic conditions. However, these simulations are non-differe…

RELATED TOPICS