AnchorComplete: Recovering Complete Scene Meshes from Partial Point Clouds

Weirong Chen, Ismail El-Yamany, Ganlin Zhang, Chuanxia Zheng, Andrea Vedaldi, Daniel CremersarXiv 2026 · arXiv

Abstract

Most image-based 3D reconstructors primarily recover visible surfaces, producing partial point clouds with large holes and leaving unobserved geometry unresolved. To address this limitation, we introduce AnchorComplete, a framework that reconstructs complete scene meshes from partial point clouds while preserving observed geometry and supporting scenes of varying scales. Building on TRELLIS.2, our method combines partial-point conditioning, which integrates global scene tokens with local spatial evidence to anchor generation to the input, and metric-scale coarse-to-fine generation, which progressively refines scene geometry within fixed-size windows while maintaining global structural consistency. Experiments on Replica and SCRREAM demonstrate that AnchorComplete consistently outperforms strong baselines in complete-scene reconstruction, particularly under sparse-view settings.
BibTeX
@misc{chen2026anchorcomplete,
  author = {Weirong Chen and Ismail El-Yamany and Ganlin Zhang and Chuanxia Zheng and Andrea Vedaldi and Daniel Cremers},
  title = {{AnchorComplete: Recovering Complete Scene Meshes from Partial Point Clouds}},
  year = {2026},
  eprint = {2611.03228},
  archivePrefix = {arXiv},
  url = {https://arxiv.org/abs/2611.03228},
}

Loading related content...

Figure