Building Rearticulable Models for Arbitrary 3D Objects from 4D Point Clouds
This addresses the challenge of modeling articulated objects for robotics and computer vision applications, representing an incremental advance over existing methods.
The paper tackles the problem of building rearticulable models for arbitrary 3D objects from 4D point clouds, identifying parts, connections, and joint properties to enable animation and retargeting to novel poses, with experiments showing it outperforms prior works on various metrics.
We build rearticulable models for arbitrary everyday man-made objects containing an arbitrary number of parts that are connected together in arbitrary ways via 1 degree-of-freedom joints. Given point cloud videos of such everyday objects, our method identifies the distinct object parts, what parts are connected to what other parts, and the properties of the joints connecting each part pair. We do this by jointly optimizing the part segmentation, transformation, and kinematics using a novel energy minimization framework. Our inferred animatable models, enables retargeting to novel poses with sparse point correspondences guidance. We test our method on a new articulating robot dataset, and the Sapiens dataset with common daily objects, as well as real-world scans. Experiments show that our method outperforms two leading prior works on various metrics.