ORGAN: Object-Centric Representation Learning using Cycle Consistent Generative Adversarial Networks

TL;DR

ORGAN introduces a novel object-centric representation learning method using cycle-consistent GANs, effectively handling complex real-world images with multiple objects and low contrast, and enabling expressive object manipulation.

Contribution

It is the first to apply cycle-consistent GANs to object-centric learning, outperforming autoencoder-based methods on challenging datasets and scaling efficiently.

Findings

Performs well on synthetic and real-world datasets

Handles many objects and low contrast images

Enables expressive object manipulation

Abstract

Although data generation is often straightforward, extracting information from data is more difficult. Object-centric representation learning can extract information from images in an unsupervised manner. It does so by segmenting an image into its subcomponents: the objects. Each object is then represented in a low-dimensional latent space that can be used for downstream processing. Object-centric representation learning is dominated by autoencoder architectures (AEs). Here, we present ORGAN, a novel approach for object-centric representation learning, which is based on cycle-consistent Generative Adversarial Networks instead. We show that it performs similarly to other state-of-the-art approaches on synthetic datasets, while at the same time being the only approach tested here capable of handling more challenging real-world datasets with many objects and low visual contrast. Complementing these results, ORGAN creates expressive latent space representations that allow for object manipulation. Finally, we show that ORGAN scales well both with respect to the number of objects and the size of the images, giving it a unique edge over current state-of-the-art approaches.