Naslov
Sim2Air - Synthetic Aerial Dataset for UAV Monitoring

Autori
Barisic, Antonella ; Petric, Frano ; Bogdan, Stjepan

Izvornik
IEEE Robotics and Automation Letters (2377-3766) 7 (2022), 2; 3757-3764

Vrsta, podvrsta i kategorija rada
Radovi u časopisima, članak, znanstveni

Ključne riječi
AI-enabled robotics, data sets for robotic vision, aerial systems: perception and autonomy

Sažetak
In this letter, we propose a novel approach to generate a synthetic aerial dataset for application in UAV monitoring. We propose to accentuate shape-based object representation by applying texture randomization. A diverse dataset with photorealism in all parameters such as shape, pose, lighting, scale, viewpoint, etc. except for atypical textures is created in a 3D modelling software Blender. Our approach specifically targets two conditions in aerial images where texture of objects is difficult to detect, namely challenging illumination and objects occupying only a small portion of the image. Experimental evaluation of YOLO and Faster R-CNN detectors trained on synthetic data with randomized textures confirmed our approach by increasing the mAP value (17 and 3.7 percentage points for YOLO ; 20 and 1.1 percentage points for Faster R-CNN) on two test datasets of real images, both containing UAV-to-UAV images with motion blur. Testing on different domains, we conclude that the more the generalisation ability is put to the test, the more apparent are the advantages of the shape-based representation.

Izvorni jezik
Engleski

Znanstvena područja
Elektrotehnika, Računarstvo, Zrakoplovstvo, raketna i svemirska tehnika

POVEZANOST RADA