HoloGAN: Unsupervised learning of 3D representations from natural images

TL;DR

HoloGAN introduces an unsupervised 3D-aware GAN that learns explicit 3D representations from natural images, enabling pose control and disentanglement of shape and appearance without supervision.

Contribution

It is the first generative model to learn 3D representations from natural images in an entirely unsupervised manner, with explicit pose control.

Findings

Enables disentanglement of 3D pose and identity

Generates high-quality images with 3D understanding

Does not require pose labels or 3D data during training

Abstract

We propose a novel generative adversarial network (GAN) for the task of unsupervised learning of 3D representations from natural images. Most generative models rely on 2D kernels to generate images and make few assumptions about the 3D world. These models therefore tend to create blurry images or artefacts in tasks that require a strong 3D understanding, such as novel-view synthesis. HoloGAN instead learns a 3D representation of the world, and to render this representation in a realistic manner. Unlike other GANs, HoloGAN provides explicit control over the pose of generated objects through rigid-body transformations of the learnt 3D features. Our experiments show that using explicit 3D features enables HoloGAN to disentangle 3D pose and identity, which is further decomposed into shape and appearance, while still being able to generate images with similar or higher visual quality than other generative models. HoloGAN can be trained end-to-end from unlabelled 2D images only. Particularly, we do not require pose labels, 3D shapes, or multiple views of the same objects. This shows that HoloGAN is the first generative model that learns 3D representations from natural images in an entirely unsupervised manner.