The goal of weapon detection is to increase public safety and help identify/ prevent dangerous situations in different environments. The way weapon detection works is to use different tools and algorithms to identify weapons quickly and accurately. This is very useful when it comes to locations like airports, schools, and other high public traffic locations. It is useful so that the public does not have to live in fear about attacks and authorities can acknowledge and address the situation in a swift manner. The three tools that are being looked at are YOLO(You Only Look Once), DETR (DEtection TRansformers), and ViT (Vision Transformer). These are various deep learning models and architectures used in computer vision and object detection tasks. The idea is to see how these tools can be used to optimize weapon detection by utilizing the best features from all three Architectures.
See the Yolo_V8 branch for all YOLO Model Results
Yolo_V8
