Learning Canonical View Representation for 3D Shape Recognition with Arbitrary Views

Open in new window