Multi-Version Analysis of YOLO: A Deep Dive into YOLOv5, YOLOv7, and YOLOv9 for Face and Person Detection

This paper compares three modern YOLO models-YOLOv5, YOLOv7, and YOLOv9-concentrating on how well they work at real-time object identification, which is a vital need for applications like driverless cars and smart surveillance. With its optimisation for edge devices, YOLOv5 provides effective performance even in contexts with limited resources. Although YOLOv9 adds multi-scale feature fusion and dynamic convolution to boost accuracy in complicated circumstances, YOLOv7 improves detection accuracy through enhanced feature pyramids. The trade-offs between detection accuracy, processing speed, and computing demands are highlighted by the experimental findings, which help choose the best YOLO model for a given set of use cases.