Rigid Body Adversarial Attacks

2025-02-08

AI Summary

This work exposes a critical limitation of the rigid-body assumption in physics simulation: finite material stiffness in reality causes trajectories from rigid-body simulators to diverge significantly from those produced by high-fidelity deformable simulators. To address this, we introduce— for the first time—the concept of adversarial attacks into physics simulation, proposing a method to synthesize “perceptually rigid” adversarial objects. These objects strictly match the appearance, geometry, and mass moment of inertia of the original rigid body, behave identically in rigid-body simulators (e.g., Bullet), yet induce trajectory deviations several times larger than reference motion scales in deformable simulators (e.g., Flex). Our approach employs gradient-based optimization to jointly model collision geometry, inertial constraints, and response discrepancies across both simulator types. We validate its effectiveness across multiple commercial simulation platforms. This work establishes a novel paradigm for quantifying the validity boundary of rigid-body approximations and enhancing simulation robustness.

Technology Category

• Computer Vision: Adversarial Attacks & Robustness
• Multiagent Systems: Adversarial Agents
• Machine Learning: Adversarial Learning & Robustness

Application Category

• User Modeling, Personalization and Recommendation: User modeling and simulation for interactive and conversational systems
• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking
• Web Mining and Content Analysis: Web data generation and simulation

Abstract

Due to their performance and simplicity, rigid body simulators are often used in applications where the objects of interest can considered very stiff. However, no material has infinite stiffness, which means there are potentially cases where the non-zero compliance of the seemingly rigid object can cause a significant difference between its trajectories when simulated in a rigid body or deformable simulator. Similarly to how adversarial attacks are developed against image classifiers, we propose an adversarial attack against rigid body simulators. In this adversarial attack, we solve an optimization problem to construct perceptually rigid adversarial objects that have the same collision geometry and moments of mass to a reference object, so that they behave identically in rigid body simulations but maximally different in more accurate deformable simulations. We demonstrate the validity of our method by comparing simulations of several examples in commercially available simulators.

Problem

• Adversarial attacks on rigid body simulators
• Constructing perceptually rigid adversarial objects
• Comparing rigid and deformable simulation behaviors

Innovation

• rigid body adversarial attacks
• optimization for identical collision geometry
• comparison in deformable simulations