Exploring the Value of Synthetic Data: Part 1

Here is a look at early results from CosmiQ Works on the impact of synthetic data on artificial intelligence models. We were proud to generate the synthetic data for the study.

This blog is the start of a two-part conclusion to our experiments with synthetic data in the RarePlanes research study. In these posts we explore the value of synthetic data from AI.Reverieand test how it could improve our ability to detect aircraft and their features. In particular we were interested in improving performance for rarely observed aircraft with limited observations in the training dataset. In our previous post, we built several baseline models using only real data from Maxar’s WorldView-3 satellite and hand annotated aircraft labels. We take the lessons learned from this post and the baselines trained only on real data and begin to investigate the best ways to apply synthetic data to augment our training dataset and improve performance in our classes of interest. Ultimately, our experiments in these two posts will test the ability to improve detection of an aircraft’s specific make and model (e.g. Cessna A-37 or Douglas C-47 Skytrain) using synthetic data.

Read more: RarePlanes — Exploring the Value of Synthetic Data: Part 1