Bridging Scene Generation and Planning: Driving with World Model via Unifying Vision and Motion Representation | Signal Canvas | ScienceToStartup