Kinematic Kitbashing

arXiv cs.RO / 5/5/2026

💬 OpinionIdeas & Deep AnalysisModels & Research

共有:

Key Points

The paper proposes “Kinematic Kitbashing,” an optimization framework for synthesizing articulated 3D objects by assembling reusable components according to an abstract kinematic graph.
It uses an exemplar-based analogy for part placement, where each reused part is matched to a single source asset to capture how it attaches to its parent.
The method models attachment context with vector distance fields and evaluates consistency by integrating matching error across the joint’s full motion range to produce a kinematics-aware attachment energy.
To add task-level functionality without requiring gradients, it uses the attachment energy as a prior in an annealed Langevin sampling approach, supporting gradient-free optimization of black-box objectives.
Experiments show the framework’s flexibility for creating graphs from user-selected or retrieved parts, synthesizing assemblies with user-defined functionality, and retargeting articulations through graph edits.

Abstract

We introduce Kinematic Kitbashing, an optimization framework that synthesizes articulated 3D objects by assembling reusable parts conditioned on an abstract kinematic graph. Given the graph and a library of articulated parts, our method optimizes per-part similarity transformations that place, orient, and scale each component into a coherent articulated object; optional graph edits further enable novel assemblies beyond the prescribed connectivity. Central to our method is an exemplar-based analogy for part placement: each reused component is paired with a single source asset that exemplifies how it attaches to its parent. We capture this attachment context using vector distance fields and measure consistency by integrating the matching error over the joint's full motion range. This yields a kinematics-aware attachment energy that favors placements that preserve the exemplar's local attachment neighborhood throughout articulation. To incorporate task-level functionality, we use this attachment energy as a prior in an annealed Langevin sampling framework, enabling gradient-free optimization of black-box functionality objectives. We demonstrate the versatility of kinematic kitbashing across diverse applications, including instantiating kinematic graphs from user-selected or automatically retrieved parts, synthesizing assemblies with user-defined functionality, and re-targeting articulations via graph edits.