Find near duplicates on your computer

Screenshot of image clustering (each to each in a folder):