Guiding Diffusion-Based Articulated Object Generation by Partial Point Cloud Alignment and Physical Plausibility Constraints

TL;DR

This paper introduces PhysNAP, a diffusion model that generates articulated objects aligned with partial point clouds and constrained by physical plausibility, enhancing realism and consistency in generated objects.

Contribution

PhysNAP is a novel diffusion-based method that incorporates point cloud alignment and physical constraints for more realistic articulated object generation.

Findings

PhysNAP improves physical plausibility of generated objects.

PhysNAP achieves better alignment with partial point clouds.

Tradeoff observed between generative diversity and constraint adherence.

Abstract

Articulated objects are an important type of interactable objects in everyday environments. In this paper, we propose PhysNAP, a novel diffusion model-based approach for generating articulated objects that aligns them with partial point clouds and improves their physical plausibility. The model represents part shapes by signed distance functions (SDFs). We guide the reverse diffusion process using a point cloud alignment loss computed using the predicted SDFs. Additionally, we impose non-penetration and mobility constraints based on the part SDFs for guiding the model to generate more physically plausible objects. We also make our diffusion approach category-aware to further improve point cloud alignment if category information is available. We evaluate the generative ability and constraint consistency of samples generated with PhysNAP using the PartNet-Mobility dataset. We also compare it with an unguided baseline diffusion model and demonstrate that PhysNAP can improve constraint consistency and provides a tradeoff with generative ability.