Unmanned aerial vehicle based multi-person detection via deep neural network models DOI Creative Commons
Mohammed Alshehri,

Laiba Zahoor,

Yahya Alqahtani

et al.

Frontiers in Neurorobotics, Journal Year: 2025, Volume and Issue: 19

Published: April 17, 2025

Understanding human actions in complex environments is crucial for advancing applications areas such as surveillance, robotics, and autonomous systems. Identifying from UAV-recorded videos becomes more challenging the task presents unique challenges, including motion blur, dynamic background, lighting variations, varying viewpoints. The presented work develops a deep learning system that recognizes multi-person behaviors data gathered by UAVs. proposed provides higher recognition accuracy while maintaining robustness along with environmental adaptability through integration of different features neural network models. study supports wider development systems utilized complicated contexts creating intelligent UAV utilizing networks. uses feature extraction approaches to create novel method recognize various video. model improves identification capacities addressing problems intricate constraints, encouraging advancements UAV-based We learning-based framework may effectively increase action scenarios. Compared existing approaches, our achieved 91.50% on MOD20 dataset 89.71% Okutama-Action. These results do, fact, show how useful network-based methods are managing limitations application. Results indeed effective at under difficult conditions.

Language: Английский

Unmanned aerial vehicle based multi-person detection via deep neural network models DOI Creative Commons
Mohammed Alshehri,

Laiba Zahoor,

Yahya Alqahtani

et al.

Frontiers in Neurorobotics, Journal Year: 2025, Volume and Issue: 19

Published: April 17, 2025

Understanding human actions in complex environments is crucial for advancing applications areas such as surveillance, robotics, and autonomous systems. Identifying from UAV-recorded videos becomes more challenging the task presents unique challenges, including motion blur, dynamic background, lighting variations, varying viewpoints. The presented work develops a deep learning system that recognizes multi-person behaviors data gathered by UAVs. proposed provides higher recognition accuracy while maintaining robustness along with environmental adaptability through integration of different features neural network models. study supports wider development systems utilized complicated contexts creating intelligent UAV utilizing networks. uses feature extraction approaches to create novel method recognize various video. model improves identification capacities addressing problems intricate constraints, encouraging advancements UAV-based We learning-based framework may effectively increase action scenarios. Compared existing approaches, our achieved 91.50% on MOD20 dataset 89.71% Okutama-Action. These results do, fact, show how useful network-based methods are managing limitations application. Results indeed effective at under difficult conditions.

Language: Английский

Citations

0