GeneSIS-RT: Generating Synthetic Images for training Secondary Real-world Tasks

TL;DR

GeneSIS-RT is a novel image-to-image translation method that generates realistic synthetic data from simulation and unlabeled real images, improving training for real-world tasks like segmentation and obstacle avoidance.

Contribution

It introduces a new approach for creating high-quality synthetic training data using image translation, reducing reliance on labeled real-world data.

Findings

Synthetic data trained models outperform raw simulated data.

Models trained with GeneSIS-RT data match or exceed real data performance.

Successfully trained a quadcopter for complex navigation tasks.

Abstract

We propose a novel approach for generating high-quality, synthetic data for domain-specific learning tasks, for which training data may not be readily available. We leverage recent progress in image-to-image translation to bridge the gap between simulated and real images, allowing us to generate realistic training data for real-world tasks using only unlabeled real-world images and a simulation. GeneSIS-RT ameliorates the burden of having to collect labeled real-world images and is a promising candidate for generating high-quality, domain-specific, synthetic data. To show the effectiveness of using GeneSIS-RT to create training data, we study two tasks: semantic segmentation and reactive obstacle avoidance. We demonstrate that learning algorithms trained using data generated by GeneSIS-RT make high-accuracy predictions and outperform systems trained on raw simulated data alone, and as well or better than those trained on real data. Finally, we use our data to train a quadcopter to fly 60 meters at speeds up to 3.4 m/s through a cluttered environment, demonstrating that our GeneSIS-RT images can be used to learn to perform mission-critical tasks.