Full-Glow: Fully conditional Glow for more realistic image generation

TL;DR

Full-Glow is a fully conditional generative model that produces realistic street scene images conditioned on semantic maps, improving data augmentation for autonomous vehicle training.

Contribution

We introduce Full-Glow, a novel conditional Glow architecture that generates more realistic images conditioned on scene layouts, enhancing training data quality for autonomous systems.

Findings

Outperforms recent models in semantic segmentation accuracy

Generated images are more similar to real scenes

Improves training data quality for autonomous vehicle perception

Abstract

Autonomous agents, such as driverless cars, require large amounts of labeled visual data for their training. A viable approach for acquiring such data is training a generative model with collected real data, and then augmenting the collected real dataset with synthetic images from the model, generated with control of the scene layout and ground truth labeling. In this paper we propose Full-Glow, a fully conditional Glow-based architecture for generating plausible and realistic images of novel street scenes given a semantic segmentation map indicating the scene layout. Benchmark comparisons show our model to outperform recent works in terms of the semantic segmentation performance of a pretrained PSPNet. This indicates that images from our model are, to a higher degree than from other models, similar to real images of the same kinds of scenes and objects, making them suitable as training data for a visual semantic segmentation or object recognition system.