Domain Stylization: A Strong, Simple Baseline for Synthetic to Real Image Domain Adaptation

TL;DR

This paper introduces a simple yet effective style transfer method for synthetic-to-real image domain adaptation, outperforming complex GAN-based approaches in semantic segmentation and object detection tasks.

Contribution

A straightforward modification to existing style transfer algorithms provides a strong baseline for synthetic-to-real domain adaptation, surpassing current state-of-the-art methods.

Findings

Achieves superior performance in semantic segmentation and object detection

Reduces Frechet Inception Distance significantly

Outperforms GAN-based image translation methods

Abstract

Deep neural networks have largely failed to effectively utilize synthetic data when applied to real images due to the covariate shift problem. In this paper, we show that by applying a straightforward modification to an existing photorealistic style transfer algorithm, we achieve state-of-the-art synthetic-to-real domain adaptation results. We conduct extensive experimental validations on four synthetic-to-real tasks for semantic segmentation and object detection, and show that our approach exceeds the performance of any current state-of-the-art GAN-based image translation approach as measured by segmentation and object detection metrics. Furthermore we offer a distance based analysis of our method which shows a dramatic reduction in Frechet Inception distance between the source and target domains, offering a quantitative metric that demonstrates the effectiveness of our algorithm in bridging the synthetic-to-real gap.