Can Synthetic Data Improve Object Detection Results for Remote Sensing Images?

TL;DR

This paper explores the use of realistic synthetic data, generated with varied parameters and refined with CycleGAN, to enhance remote sensing object detection accuracy, especially when real data is limited.

Contribution

It introduces a novel approach combining synthetic data generation, pixel-level refinement with CycleGAN, and fine-tuning to improve remote sensing object detection performance.

Findings

Synthetic data improves detection accuracy on multiple datasets.

Refinement with CycleGAN enhances realism of synthetic images.

Combining synthetic and real data yields better results than using real data alone.

Abstract

Deep learning approaches require enough training samples to perform well, but it is a challenge to collect enough real training data and label them manually. In this letter, we propose the use of realistic synthetic data with a wide distribution to improve the performance of remote sensing image aircraft detection. Specifically, to increase the variability of synthetic data, we randomly set the parameters during rendering, such as the size of the instance and the class of background images. In order to make the synthetic images more realistic, we then refine the synthetic images at the pixel level using CycleGAN with real unlabeled images. We also fine-tune the model with a small amount of real data, to obtain a higher accuracy. Experiments on NWPU VHR-10, UCAS-AOD and DIOR datasets demonstrate that the proposed method can be applied for augmenting insufficient real data.