CRM: Single Image to 3D Textured Mesh with Convolutional Reconstruction Model

TL;DR

CRM is a fast, high-fidelity single-image 3D reconstruction model that integrates geometric priors and convolutional architectures to produce textured meshes in seconds without test-time optimization.

Contribution

The paper introduces CRM, a novel convolutional-based model that effectively incorporates geometric priors for high-quality 3D mesh generation from a single image.

Findings

Generates textured 3D meshes in 10 seconds without optimization.

Leverages spatial correspondence of orthographic views for better quality.

Employs Flexicubes for direct end-to-end textured mesh optimization.

Abstract

Feed-forward 3D generative models like the Large Reconstruction Model (LRM) have demonstrated exceptional generation speed. However, the transformer-based methods do not leverage the geometric priors of the triplane component in their architecture, often leading to sub-optimal quality given the limited size of 3D data and slow training. In this work, we present the Convolutional Reconstruction Model (CRM), a high-fidelity feed-forward single image-to-3D generative model. Recognizing the limitations posed by sparse 3D data, we highlight the necessity of integrating geometric priors into network design. CRM builds on the key observation that the visualization of triplane exhibits spatial correspondence of six orthographic images. First, it generates six orthographic view images from a single input image, then feeds these images into a convolutional U-Net, leveraging its strong pixel-level alignment capabilities and significant bandwidth to create a high-resolution triplane. CRM further employs Flexicubes as geometric representation, facilitating direct end-to-end optimization on textured meshes. Overall, our model delivers a high-fidelity textured mesh from an image in just 10 seconds, without any test-time optimization.