Two-stage detectors: R-CNN to Faster R-CNN · MAAI7602 Applied Computer Vision

Two-stage detectors first propose candidate regions and then classify and refine each one. Each generation removed repeated work: shared features (Fast R-CNN) and learned proposals (Faster R-CNN). How do two-stage detectors propose and classify regions, and how did R-CNN, Fast R-CNN and Faster R-CNN reduce their cost?

This page needs JavaScript for its explanations, activities and practice questions.