# What Do Single-view 3D Reconstruction Networks Learn?

## TL;DR

This paper critically examines single-view 3D reconstruction networks, revealing that current methods mainly perform image classification rather than true 3D reconstruction, and proposes simpler baselines that outperform existing techniques.

## Contribution

It introduces simple classification and retrieval baselines that surpass state-of-the-art reconstruction methods and questions the actual reconstruction capabilities of current models.

## Key findings

- Baseline methods outperform existing reconstruction techniques.
- Encoder-decoder models are statistically similar to simple baselines.
- Current methods mainly perform image classification, not true 3D reconstruction.

## Abstract

Convolutional networks for single-view object reconstruction have shown impressive performance and have become a popular subject of research. All existing techniques are united by the idea of having an encoder-decoder network that performs non-trivial reasoning about the 3D structure of the output space. In this work, we set up two alternative approaches that perform image classification and retrieval respectively. These simple baselines yield better results than state-of-the-art methods, both qualitatively and quantitatively. We show that encoder-decoder methods are statistically indistinguishable from these baselines, thus indicating that the current state of the art in single-view object reconstruction does not actually perform reconstruction but image classification. We identify aspects of popular experimental procedures that elicit this behavior and discuss ways to improve the current state of research.