Face recognition in drone-shot videos

Abstract

Unmanned Aerial Vehicles (UAVs) or drones are often used to reach remote areas or regions which are inaccessible to humans. Equipped with a large eld of view, compact size, and remote control abilities, drones are deemed suitable for monitoring crowded or disaster-hit areas and performing aerial surveillance. While research has focused on area monitoring, object detection, and tracking, limited attention has been given to person identi cation, especially face recognition, using drones. In this research, we present the DroneSURF dataset [12], a novel large-scale drone videos dataset, collected over two segments. DroneSURF dataset is intended to facilitate research for drone-based face detection and recognition. For both segments, we provide videos, high-resolution gallery images, and annotated face regions. In addition to that, for segment-1 of DroneSURF, we also provide baseline results on face detection and face recognition and the results of our own experiments to improve the performance of face recognition in drone-shot videos.