Deep Convolutional Inverse Graphics Network

TL;DR

The paper introduces DC-IGN, a deep convolutional network that learns disentangled, interpretable representations of images, enabling controlled generation of images with variations in pose and lighting.

Contribution

It proposes a novel training method for disentangling transformations in image representations using convolutional networks and variational Bayes.

Findings

Successfully learns interpretable representations of pose and lighting.

Can generate images with controlled variations in pose and lighting.

Demonstrates effectiveness as a 3D rendering engine.

Abstract

This paper presents the Deep Convolution Inverse Graphics Network (DC-IGN), a model that learns an interpretable representation of images. This representation is disentangled with respect to transformations such as out-of-plane rotations and lighting variations. The DC-IGN model is composed of multiple layers of convolution and de-convolution operators and is trained using the Stochastic Gradient Variational Bayes (SGVB) algorithm. We propose a training procedure to encourage neurons in the graphics code layer to represent a specific transformation (e.g. pose or light). Given a single input image, our model can generate new images of the same object with variations in pose and lighting. We present qualitative and quantitative results of the model's efficacy at learning a 3D rendering engine.