Fujitsu and Carnegie Mellon University have developed a new technology to visualize traffic situations, including people and vehicles. The technology transforms a 2D scene image captured by a monocular RGB camera into a digitalized 3D format using AI. This technology relies on AI that has been trained to detect the shape of people and objects through deep learning. By utilizing these technologies, images taken in situations in which people and cars are densely situated, such as at intersections, can be dynamically reconstructed