2019/05/16 by Mohamed Afifi, Afifi, Mohamed, Yara Ali +8 · 6 citations
Computer Science · Engineering · #Advanced Neural Network Applications #Aerial image #Artificial intelligence #Autonomous Vehicle Technology and Safety #Computer Vision and Pattern Recognition (cs.CV) #Computer science #Computer security #Computer vision #Deep learning #Drone #Engineering #Exploit #FOS: Computer and information sciences #Frame (networking) #Geography #Image (mathematics) #Intersection (aeronautics) #Object detection #Patrolling #Pattern recognition (psychology) #Pedestrian #Pedestrian detection #Real-time computing #Throughput #Video Surveillance and Tracking Methods #cs.CV
paper · pdf · doi:10.48550/arxiv.1905.06653
published in arXiv (Cornell University) (Cornell University)
arxiv created 2019/05/16 · openalex publication_date 2019/05/16 · arxiv updated 2019/05/17 · openalex created_date 2025/10/10 · openalex updated_date 2026/07/28
Detection of pedestrians in aerial imagery captured by drones has many applications including intersection monitoring, patrolling, and surveillance, to name a few. However, the problem is involved due to continuouslychanging camera viewpoint and object appearance as well as the need for lightweight algorithms to run on on-board embedded systems. To address this issue, the paper proposes a framework for pedestrian detection in videos based on the YOLO object detection network [6] while having a high throughput of more than 5 FPS on the Jetson TX2 embedded board. The framework exploits deep learning for robust operation and uses a pre-trained model without the need for any additional training which makes it flexible to apply on different setups with minimum amount of tuning. The method achieves ~81 mAP when applied on a sample video from the Embedded Real-Time Inference (ERTI) Challenge where pedestrians are monitored by a UAV.