1 paper
Tom Monnier, Jake Austin, Angjoo Kanazawa +2
Given a set of calibrated images of a scene, we present an approach that produces a simple, compact, and actionable 3D world representation by means of 3D primitives. While many ap…