Kinematify: Open-Vocabulary Synthesis of High-DoF Articulated Objects

TL;DR

Kinematify is an automated framework that synthesizes high-DoF articulated object models from RGB images or text, combining search and optimization to infer structure and joint parameters for robotics applications.

Contribution

It introduces a novel method that infers kinematic structures and joint parameters of high-DoF objects directly from static images or descriptions, improving scalability and accuracy.

Findings

Enhanced kinematic topology accuracy over prior methods

Effective synthesis from both synthetic and real-world data

Improved registration and joint parameter estimation

Abstract

A deep understanding of kinematic structures and movable components is essential for enabling robots to manipulate objects and model their own articulated forms. Such understanding is captured through articulated objects, which are essential for tasks such as physical simulation, motion planning, and policy learning. However, creating these models, particularly for objects with high degrees of freedom (DoF), remains a significant challenge. Existing methods typically rely on motion sequences or strong assumptions from hand-curated datasets, which hinders scalability. In this paper, we introduce Kinematify, an automated framework that synthesizes articulated objects directly from arbitrary RGB images or textual descriptions. Our method addresses two core challenges: (i) inferring kinematic topologies for high-DoF objects and (ii) estimating joint parameters from static geometry. To achieve this, we combine MCTS search for structural inference with geometry-driven optimization for joint reasoning, producing physically consistent and functionally valid descriptions. We evaluate Kinematify on diverse inputs from both synthetic and real-world environments, demonstrating improvements in registration and kinematic topology accuracy over prior work.