Oriel — AI & technology worth noticing, every day.

ObjectTwin: Measuring the Asset Correspondence Gap for Capture-Anchored Digital Twins

Paper · https://arxiv.org/abs/2610.35904

arXiv:2610.35904v1 Announce Type: new Abstract: Object-wise reconstruction increasingly inserts generated meshes into captured scenes as editable proxies for observed objects. This use creates a requirement absent from isolated-asset scores and pose estimates: at the captured metric state, the proxy should reproduce calibrated observations of the same instance. We present ObjectTwin, a scan-grounded measurement protocol that freezes the instance, cameras, scan-projected support and recorded depth, metric anchor, and evaluation denominator, then replaces only the object representation. The resulting asset correspondence gap measures lost observation agreement relative to scan-derived visible geometry without treating the scan as complete physical ground truth. Across two indoor RGB-D capture domains and two generation backends, visible, anchored generated assets lose 0.302-0.497 MaskIoU under this replacement. Every tested scene-level estimate is positive, while same-input generator preference reverses across capture domains. Coverage and overlap-depth analyses locate disagreement in both visible support and surface placement; the evaluated conservative pose refinement does not systematically recover it. These results establish capture correspondence as a separately measurable interface between generated-asset quality and stronger physical or task-level evaluation.

First seen by Oriel on .

AI summary

ObjectTwin introduces a scan-grounded protocol for measuring how well generated object proxies reproduce calibrated observations of specific captured instances, finding substantial correspondence gaps across tested domains and generators.

  • The protocol fixes the instance, cameras, scan-projected support, recorded depth, metric anchor, and evaluation denominator, changing only the object representation.
  • Across two indoor RGB-D capture domains and two generation backends, generated assets lose 0.302–0.497 MaskIoU against scan-derived visible geometry.
  • All tested scene-level estimates are positive, but preference between generators using the same input reverses across capture domains.
  • Disagreement reflects both visible-support and surface-placement errors; conservative pose refinement does not systematically recover it.
  • The authors position capture correspondence as a distinct evaluation interface, separate from generated-asset quality and stronger physical or task-level evaluation.

#digital twins #3d reconstruction #rgb-d #computer vision

Generated by gpt-5.6-sol. Check the original for details.

Aliases

SourceExternal IDURL
arXiv AI Papersoai:arXiv.org:2610.35904v1https://arxiv.org/abs/2610.35904 Read discussion