Synthetic Data for Object Classification in Industrial Applications

TL;DR

This paper investigates using synthetic images generated by a game engine combined with limited real data to effectively train object classifiers for industrial applications, reducing data collection efforts.

Contribution

It introduces a hybrid training approach that leverages synthetic data and minimal real images, achieving high accuracy with significantly less real data needed.

Findings

Synthetic data improves classifier confidence.

High accuracy achieved with only 12-24 real images per class.

Combining synthetic and real data reduces data collection efforts.

Abstract

One of the biggest challenges in machine learning is data collection. Training data is an important part since it determines how the model will behave. In object classification, capturing a large number of images per object and in different conditions is not always possible and can be very time-consuming and tedious. Accordingly, this work explores the creation of artificial images using a game engine to cope with limited data in the training dataset. We combine real and synthetic data to train the object classification engine, a strategy that has shown to be beneficial to increase confidence in the decisions made by the classifier, which is often critical in industrial setups. To combine real and synthetic data, we first train the classifier on a massive amount of synthetic data, and then we fine-tune it on real images. Another important result is that the amount of real images needed for fine-tuning is not very high, reaching top accuracy with just 12 or 24 images per class. This substantially reduces the requirements of capturing a great amount of real data.