$\mathbf{M^3A}$ Policy: Mutable Material Manipulation Augmentation Policy through Photometric Re-rendering

TL;DR

This paper introduces M$^3$A, a photometric re-rendering framework that generates diverse material appearances from a single demonstration, significantly improving robotic manipulation policies' ability to generalize across different materials.

Contribution

The paper presents a novel material augmentation method using photometric re-rendering to enhance cross-material generalization in robotic manipulation tasks.

Findings

Improves success rate across three real-world tasks by 58.03%.

Enhances robustness on unseen materials.

Constructs the first multi-material manipulation benchmark.

Abstract

Material generalization is essential for real-world robotic manipulation, where robots must interact with objects exhibiting diverse visual and physical properties. This challenge is particularly pronounced for objects made of glass, metal, or other materials whose transparent or reflective surfaces introduce severe out-of-distribution variations. Existing approaches either rely on simulated materials in simulators and perform sim-to-real transfer, which is hindered by substantial visual domain gaps, or depend on collecting extensive real-world demonstrations, which is costly, time-consuming, and still insufficient to cover various materials. To overcome these limitations, we resort to computational photography and introduce Mutable Material Manipulation Augmentation (M$^3$A), a unified framework that leverages the physical characteristics of materials as captured by light transport for photometric re-rendering. The core idea is simple yet powerful: given a single real-world demonstration, we photometrically re-render the scene to generate a diverse set of highly realistic demonstrations with different material properties. This augmentation effectively decouples task-specific manipulation skills from surface appearance, enabling policies to generalize across materials without additional data collection. To systematically evaluate this capability, we construct the first comprehensive multi-material manipulation benchmark spanning both simulation and real-world environments. Extensive experiments show that the M$^3$A policy significantly enhances cross-material generalization, improving the average success rate across three real-world tasks by 58.03\%, and demonstrating robust performance on previously unseen materials.