We focus on building a tool that helps uncover useful clues from public images, like signs, text, places, and landmarks.