SuperMat: Physically Consistent PBR Material Estimation at Interactive Rates

TL;DR

SuperMat is a fast, physically consistent PBR material estimation method that decomposes material properties from images in milliseconds and extends to 3D objects with viewpoint consistency.

Contribution

It introduces a single-step, end-to-end framework for high-quality material decomposition that is significantly faster than previous diffusion-based methods.

Findings

Achieves state-of-the-art decomposition quality.

Reduces inference time from seconds to milliseconds.

Estimates materials for 3D objects in about 3 seconds.

Abstract

Decomposing physically-based materials from images into their constituent properties remains challenging, particularly when maintaining both computational efficiency and physical consistency. While recent diffusion-based approaches have shown promise, they face substantial computational overhead due to multiple denoising steps and separate models for different material properties. We present SuperMat, a single-step framework that achieves high-quality material decomposition with one-step inference. This enables end-to-end training with perceptual and re-render losses while decomposing albedo, metallic, and roughness maps at millisecond-scale speeds. We further extend our framework to 3D objects through a UV refinement network, enabling consistent material estimation across viewpoints while maintaining efficiency. Experiments demonstrate that SuperMat achieves state-of-the-art PBR material decomposition quality while reducing inference time from seconds to milliseconds per image, and completes PBR material estimation for 3D objects in approximately 3 seconds. The project page is at https://hyj542682306.github.io/SuperMat/.