What is YOLO? The Ultimate Guide [2025]

YOLO, standing for "You Only Look Once," is a family of computer vision models that revolutionized real-time object detection since its introduction by Joseph Redmon and colleagues in 2016. Originating from a custom framework called Darknet, YOLO models are single-stage detectors that efficiently combine object identification and classification in one pass, making them faster and more suitable for applications requiring real-time processing compared to traditional two-stage detectors like Faster R-CNN. Over the years, the YOLO family has expanded with contributions from various researchers, resulting in multiple versions, each with unique enhancements aimed at improving speed, accuracy, and ease of deployment on edge devices. These models have been benchmarked using the Microsoft COCO dataset, achieving notable performance in terms of mean Average Precision (mAP). YOLO's open-source nature has fostered a large community, facilitating widespread adoption and continuous innovation, with practical applications ranging from monitoring traffic patterns to detecting safety violations in industrial settings.