SurfNet: Generating 3D shape surfaces using deep residual networks

TL;DR

This paper introduces a deep residual network approach for directly generating 3D shape surfaces using geometry images, enabling shape interpolation, pose variation, and reconstruction from images, bypassing voxel-based methods.

Contribution

The authors propose a novel method to generate 3D shape surfaces directly with deep residual networks using geometry images, improving efficiency and detail over voxel-based approaches.

Findings

Network learns meaningful shape surface representations

Enables interpolation between shapes and poses

Reconstructs 3D surfaces from unseen images

Abstract

3D shape models are naturally parameterized using vertices and faces, \ie, composed of polygons forming a surface. However, current 3D learning paradigms for predictive and generative tasks using convolutional neural networks focus on a voxelized representation of the object. Lifting convolution operators from the traditional 2D to 3D results in high computational overhead with little additional benefit as most of the geometry information is contained on the surface boundary. Here we study the problem of directly generating the 3D shape surface of rigid and non-rigid shapes using deep convolutional neural networks. We develop a procedure to create consistent `geometry images' representing the shape surface of a category of 3D objects. We then use this consistent representation for category-specific shape surface generation from a parametric representation or an image by developing novel extensions of deep residual networks for the task of geometry image generation. Our experiments indicate that our network learns a meaningful representation of shape surfaces allowing it to interpolate between shape orientations and poses, invent new shape surfaces and reconstruct 3D shape surfaces from previously unseen images.