A deep learning model for firearm detection in surveillance videos

This work focuses on the detection of firearms in non-canonical perspectives in surveillance videos using deep learning with a focus on computational efficiency. In particular, the performance of YOLOv8, a state-of-the-art architecture, is evaluated using the nano version, its most efficient variant...

• Glavni autor: Crespo Yaguana, Brian Anthony (author)
• Format: bachelorThesis
• Jezik: eng
• Izdano: 2024
• Teme: Detección de objetos; Visión artificial; Aprendizaje profundo; Object detection; Computer vision; Deep learning
• Online pristup: http://repositorio.yachaytech.edu.ec/handle/123456789/863