Visionary vigilance: Optimized YOLOV8 for fallen person detection with large-scale benchmark dataset

Open Webpage

Habib Khan, Inam Ullah, Mohammad Shabaz, Muhammad Faizan Omer, Muhammad Talha Usman, Mohammed Seghir Guellil, Jakeoung Koo

Published: 01 Jan 2024, Last Modified: 24 Jul 2025Image Vis. Comput. 2024EveryoneRevisionsBibTeXCC BY-SA 4.0
Abstract: Highlights•Developed DiverseFALL10500: a comprehensive dataset with 10,500 annotated images.•Optimized YOLOv8s with focus module and CBAM integration for improved performance.•Demonstrated superiority over SOTA techniques through extensive empirical evaluation.