YOLO (You Only Look Once) is a cutting-edge real-time object detection system known for its speed and accuracy. YOLOv3 processes images at 30 FPS with a mAP of 57.9% on COCO test-dev. It offers exceptional efficiency, being on par with Focal Loss but about four times faster. YOLO's unique feature allows easy trade-offs between speed and accuracy by adjusting the model size without retraining.
Unlike previous systems, YOLO uses a single neural network for the entire image, making predictions informed by global context and achieving remarkable speed—over 1000 times faster than R-CNN and 100 times faster than Fast R-CNN. Its ability to make predictions with just one network evaluation sets it apart from systems requiring thousands for a single image.
