ArtGen: Conditional Generative Modeling of Articulated Objects in Arbitrary Part-Level States
This addresses the challenge of generating articulated assets for robotics, digital twins, and embodied intelligence, representing a novel method for a known bottleneck.
The paper tackled the problem of generating articulated 3D objects with ambiguous or unrealistic kinematic structures from single-view inputs by introducing ArtGen, a conditional diffusion-based framework that uses cross-state Monte Carlo sampling and a Chain-of-Thought reasoning module, resulting in significant outperformance over state-of-the-art methods on the PartNet-Mobility benchmark.
Generating articulated assets is crucial for robotics, digital twins, and embodied intelligence. Existing generative models often rely on single-view inputs representing closed states, resulting in ambiguous or unrealistic kinematic structures due to the entanglement between geometric shape and joint dynamics. To address these challenges, we introduce ArtGen, a conditional diffusion-based framework capable of generating articulated 3D objects with accurate geometry and coherent kinematics from single-view images or text descriptions at arbitrary part-level states. Specifically, ArtGen employs cross-state Monte Carlo sampling to explicitly enforce global kinematic consistency, reducing structural-motion entanglement. Additionally, we integrate a Chain-of-Thought reasoning module to infer robust structural priors, such as part semantics, joint types, and connectivity, guiding a sparse-expert Diffusion Transformer to specialize in diverse kinematic interactions. Furthermore, a compositional 3D-VAE latent prior enhanced with local-global attention effectively captures fine-grained geometry and global part-level relationships. Extensive experiments on the PartNet-Mobility benchmark demonstrate that ArtGen significantly outperforms state-of-the-art methods.