Improved Modeling of 3D Shapes with Multi-view Depth Maps

TL;DR

This paper introduces a versatile framework that uses 2D CNN techniques to generate dense multi-view depth maps from a single depth image, enabling efficient and high-quality 3D shape modeling.

Contribution

It proposes a novel encoder-decoder architecture with an identity encoder and class-conditional viewpoint generator for 3D shape reconstruction.

Findings

Outperforms existing depth map methods in 3D reconstruction quality.

Enables high-resolution 3D shape generation with low memory usage.

Compatible with architectures from 2D image domain for 3D modeling.

Abstract

We present a simple yet effective general-purpose framework for modeling 3D shapes by leveraging recent advances in 2D image generation using CNNs. Using just a single depth image of the object, we can output a dense multi-view depth map representation of 3D objects. Our simple encoder-decoder framework, comprised of a novel identity encoder and class-conditional viewpoint generator, generates 3D consistent depth maps. Our experimental results demonstrate the two-fold advantage of our approach. First, we can directly borrow architectures that work well in the 2D image domain to 3D. Second, we can effectively generate high-resolution 3D shapes with low computational memory. Our quantitative evaluations show that our method is superior to existing depth map methods for reconstructing and synthesizing 3D objects and is competitive with other representations, such as point clouds, voxel grids, and implicit functions.