Synthetic Image Data for Deep Learning

TL;DR

This paper demonstrates that high-quality synthetic images, generated through physically-based rendering and domain randomization, can effectively augment real datasets for training deep learning models in vehicle image classification and segmentation, reducing training costs.

Contribution

It introduces a method for creating large synthetic datasets from 3D CAD models and quantifies their effectiveness in improving model accuracy and training efficiency.

Findings

Synthetic images improve segmentation accuracy when combined with real data.

Adding even small amounts of real data to synthetic datasets greatly enhances performance.

Pretraining on synthetic data significantly reduces transfer learning training time.

Abstract

Realistic synthetic image data rendered from 3D models can be used to augment image sets and train image classification semantic segmentation models. In this work, we explore how high quality physically-based rendering and domain randomization can efficiently create a large synthetic dataset based on production 3D CAD models of a real vehicle. We use this dataset to quantify the effectiveness of synthetic augmentation using U-net and Double-U-net models. We found that, for this domain, synthetic images were an effective technique for augmenting limited sets of real training data. We observed that models trained on purely synthetic images had a very low mean prediction IoU on real validation images. We also observed that adding even very small amounts of real images to a synthetic dataset greatly improved accuracy, and that models trained on datasets augmented with synthetic images were more accurate than those trained on real images alone. Finally, we found that in use cases that benefit from incremental training or model specialization, pretraining a base model on synthetic images provided a sizeable reduction in the training cost of transfer learning, allowing up to 90\% of the model training to be front-loaded.