SynDrone -- Multi-modal UAV Dataset for Urban Scenarios

TL;DR

This paper introduces SynDrone, a large-scale multimodal synthetic UAV dataset with high-resolution images, 3D data, and detailed annotations across 28 classes, aimed at advancing computer vision for urban UAV scenarios.

Contribution

The paper presents a novel synthetic dataset with multimodal data and pixel-level annotations, facilitating improved training and evaluation of deep learning models for UAV applications.

Findings

Effective training of deep architectures demonstrated

Promising results in synthetic-to-real domain adaptation

Dataset supports semantic segmentation and object detection

Abstract

The development of computer vision algorithms for Unmanned Aerial Vehicles (UAVs) imagery heavily relies on the availability of annotated high-resolution aerial data. However, the scarcity of large-scale real datasets with pixel-level annotations poses a significant challenge to researchers as the limited number of images in existing datasets hinders the effectiveness of deep learning models that require a large amount of training data. In this paper, we propose a multimodal synthetic dataset containing both images and 3D data taken at multiple flying heights to address these limitations. In addition to object-level annotations, the provided data also include pixel-level labeling in 28 classes, enabling exploration of the potential advantages in tasks like semantic segmentation. In total, our dataset contains 72k labeled samples that allow for effective training of deep architectures showing promising results in synthetic-to-real adaptation. The dataset will be made publicly available to support the development of novel computer vision methods targeting UAV applications.