Attention-based 3D Object Reconstruction from a Single Image

TL;DR

This paper introduces an attention-based enhancement to 3D object reconstruction from a single image, improving global feature extraction and achieving higher accuracy over existing methods.

Contribution

It applies self-attention within the encoder of Occupancy Networks, significantly boosting reconstruction quality and mesh consistency compared to prior convolutional approaches.

Findings

5.05% improvement in mesh IoU

0.83% increase in Normal Consistency

Over 10X reduction in Chamfer-L1 distance

Abstract

Recently, learning-based approaches for 3D reconstruction from 2D images have gained popularity due to its modern applications, e.g., 3D printers, autonomous robots, self-driving cars, virtual reality, and augmented reality. The computer vision community has applied a great effort in developing functions to reconstruct the full 3D geometry of objects and scenes. However, to extract image features, they rely on convolutional neural networks, which are ineffective in capturing long-range dependencies. In this paper, we propose to substantially improve Occupancy Networks, a state-of-the-art method for 3D object reconstruction. For such we apply the concept of self-attention within the network's encoder in order to leverage complementary input features rather than those based on local regions, helping the encoder to extract global information. With our approach, we were capable of improving the original work in 5.05% of mesh IoU, 0.83% of Normal Consistency, and more than 10X the Chamfer-L1 distance. We also perform a qualitative study that shows that our approach was able to generate much more consistent meshes, confirming its increased generalization power over the current state-of-the-art.