In simple words: YOLO is an object detector designed for fast predictions.
One main pass
You Only Look Once (YOLO) predicts object locations and classes from an image in one main model pass. This design is useful when speed matters, such as processing video frames.
Accuracy and speed
Real systems compare speed and accuracy on their own data. YOLO and region-based methods are different design families, and no one method is best for every image or device.
See the flow
video frame→YOLO→labelled bounding boxes
Simple example
A warehouse camera can locate parcels and people in each frame so staff can monitor a process.
Remember: YOLO is an object detector designed for fast predictions.