DM-OSVP++: One-Shot View Planning Using 3D Diffusion Models for Active RGB-Based Object Reconstruction




Authors:

S. Pan, L. Jin, X. Huang, C. Stachniss, M. Popović, M. Bennewitz

Type:

Preprint

Published in:

Arxiv Pre-print

Year:

2025

Related Projects:

AID4Crops - Automation and AI for Monitoring and Decision Making of Horticultural Crops, Phenorob - Robotics and Phenotyping for Sustainable Crop Production, RePAIR - Reconstructing the Past: Artificial Intelligence and Robotics Meet Cultural Heritage

Links:

Preprint

BibTex String

@article{pan2025dmosvp,
title={DM-OSVP++: One-Shot View Planning Using 3D Diffusion Models for Active RGB-Based Object Reconstruction},
author={Pan, Sicong and Jin, Liren and Huang, Xuying and Stachniss, Cyrill and Popović, Marija and Bennewitz, Maren},
journal={arXiv preprint arXiv:2504.11674},
year={2025}
}
Topic

Abstract:

Active object reconstruction is crucial for many robotic applications. A key aspect in these scenarios is generating object-specific view configurations to obtain informative measurements for reconstruction. One-shot view planning enables efficient data collection by predicting all views at once, eliminating the need for time-consuming online replanning. Our primary insight is to leverage the generative power of 3D diffusion models as valuable prior information. By conditioning on initial multi-view images, we exploit the priors from the 3D diffusion model to generate an approximate object model, serving as the foundation for our view planning. Our novel approach integrates the geometric and textural distributions of the object model into the view planning process, generating views that focus on the complex parts of the object to be reconstructed. We validate the proposed active object reconstruction system through both simulation and real-world experiments, demonstrating the effectiveness of using 3D diffusion priors for one-shot view planning.