Object detection algorithms fall into two main groups. (1) Classification-based (two-stage) detectors first select regions of interest in the image and then classify each region with a convolutional neural network. Because a prediction is run for every proposed region, they tend to be accurate but slower. The Region-based Convolutional Neural Network (R-CNN) and its successors—Fast R-CNN, Faster R-CNN, and Mask R-CNN—are well-known examples. (2) Regression-based (one-stage) detectors skip the region-proposal step and instead predict object classes and bounding boxes for the whole image in a single pass. They trade a little accuracy for large gains in speed, making them well suited to real-time detection; the best-known examples are the YOLO (You Only Look Once) family and the Single Shot MultiBox Detector (SSD).