2022/01/01 by Daniel Organisciak, Matthew Poyser, Aishah Alsehaim +4 · 1 citation
Computer Science · Engineering · #Video Surveillance and Tracking Methods #Advanced Neural Network Applications #UAV Applications and Optimization #Benchmark (surveying) #Identification (biology) #Computer science #Remote sensing #Aerial imagery #Artificial intelligence #Computer vision #Remotely operated underwater vehicle #Geography #Mobile robot #Cartography #Robot
paper · doi:10.5220/0010836600003124
openalex publication_date 2022/01/01 · openalex created_date 2025/10/10 · openalex updated_date 2026/07/29
As unmanned aerial vehicles (UAV) become more accessible with a growing range of applications, \nthe risk of UAV disruption increases. Recent development in deep learning allows vision-based \ncounter-UAV systems to detect and track UAVs with a single camera. However, the limited eld of \nview of a single camera necessitates multi-camera congurations to match UAVs across viewpoints \n a problem known as re-identication (Re-ID). While there has been extensive research on person \nand vehicle Re-ID to match objects across time and viewpoints, to the best of our knowledge, \nUAV Re-ID remains unresearched but challenging due to great dierences in scale and pose. We \npropose the rst UAV re-identication data set, UAV-reID, to facilitate the development of machine \nlearning solutions in multi-camera environments. UAV-reID has two sub-challenges: Temporally- \nNear and Big-to-Small to evaluate Re-ID performance across viewpoints and scale respectively. \nWe conduct a benchmark study by extensively evaluating dierent Re-ID deep learning based \napproaches and their variants, spanning both convolutional and transformer architectures. Under \nthe optimal conguration, such approaches are suciently powerful to learn a well-performing \nrepresentation for UAV (81.9% mAP for Temporally-Near, 46.5% for the more dicult Big-to- \nSmall challenge), while vision transformers are the most robust to extreme variance of scale.