Controlling generative models with continuous factors of variations

TL;DR

This paper introduces a new method to interpret and control the latent space of generative models, enabling precise manipulation of image properties without human annotations, applicable to GANs and auto-encoders.

Contribution

The paper presents a novel, annotation-free approach to identify meaningful directions in latent spaces for controlling specific image attributes.

Findings

Effective control over image translation, zoom, and color variations

Applicable to both GANs and variational auto-encoders

Qualitative and quantitative validation of the method

Abstract

Recent deep generative models are able to provide photo-realistic images as well as visual or textual content embeddings useful to address various tasks of computer vision and natural language processing. Their usefulness is nevertheless often limited by the lack of control over the generative process or the poor understanding of the learned representation. To overcome these major issues, very recent work has shown the interest of studying the semantics of the latent space of generative models. In this paper, we propose to advance on the interpretability of the latent space of generative models by introducing a new method to find meaningful directions in the latent space of any generative model along which we can move to control precisely specific properties of the generated image like the position or scale of the object in the image. Our method does not require human annotations and is particularly well suited for the search of directions encoding simple transformations of the generated image, such as translation, zoom or color variations. We demonstrate the effectiveness of our method qualitatively and quantitatively, both for GANs and variational auto-encoders.