# Domain-Adaptive Single-View 3D Reconstruction

## TL;DR

This paper introduces a domain-adaptive framework for single-view 3D shape reconstruction that uses adversarial training to bridge the gap between synthetic and natural images and to ensure realistic reconstructions, achieving competitive results.

## Contribution

It presents a novel adversarial training approach that simultaneously reduces domain gap and enforces realistic shape constraints in 3D reconstruction from single images.

## Key findings

- Significant performance improvement over baseline models.
- Competitive results with state-of-the-art methods.
- Simpler architecture with effective domain adaptation.

## Abstract

Single-view 3D shape reconstruction is an important but challenging problem, mainly for two reasons. First, as shape annotation is very expensive to acquire, current methods rely on synthetic data, in which ground-truth 3D annotation is easy to obtain. However, this results in domain adaptation problem when applied to natural images. The second challenge is that there are multiple shapes that can explain a given 2D image. In this paper, we propose a framework to improve over these challenges using adversarial training. On one hand, we impose domain confusion between natural and synthetic image representations to reduce the distribution gap. On the other hand, we impose the reconstruction to be `realistic' by forcing it to lie on a (learned) manifold of realistic object shapes. Our experiments show that these constraints improve performance by a large margin over baseline reconstruction models. We achieve results competitive with the state of the art with a much simpler architecture.