SPLATART: Articulated Gaussian Splatting with Estimated Object Structure

TL;DR

SPLATART is a novel pipeline that learns Gaussian splat representations of articulated objects from images, effectively capturing geometry, part separation, and joint structure, even for objects with complex kinematic trees.

Contribution

It introduces a method to disentangle part separation from articulation estimation, enabling modeling of objects with deeper kinematic trees from posed images.

Findings

Successfully applied to synthetic Paris dataset objects.

Qualitative results on real-world objects with minimal supervision.

Demonstrated effectiveness on articulated manipulators with deep kinematic structures.

Abstract

Representing articulated objects remains a difficult problem within the field of robotics. Objects such as pliers, clamps, or cabinets require representations that capture not only geometry and color information, but also part seperation, connectivity, and joint parametrization. Furthermore, learning these representations becomes even more difficult with each additional degree of freedom. Complex articulated objects such as robot arms may have seven or more degrees of freedom, and the depth of their kinematic tree may be notably greater than the tools, drawers, and cabinets that are the typical subjects of articulated object research. To address these concerns, we introduce SPLATART - a pipeline for learning Gaussian splat representations of articulated objects from posed images, of which a subset contains image space part segmentations. SPLATART disentangles the part separation task from the articulation estimation task, allowing for post-facto determination of joint estimation and representation of articulated objects with deeper kinematic trees than previously exhibited. In this work, we present data on the SPLATART pipeline as applied to the syntheic Paris dataset objects, and qualitative results on a real-world object under spare segmentation supervision. We additionally present on articulated serial chain manipulators to demonstrate usage on deeper kinematic tree structures.