no code implementations • NeurIPS 2011 • Andreas Geiger, Christian Wojek, Raquel Urtasun
We propose a novel generative model that is able to reason jointly about the 3D scene layout as well as the 3D location and orientation of objects in the scene.