Revar3r: gauge-aware perturbation uncertainty for feed-forward 3d reconstruction

Key points

• This exposes a weakness of trainingfree perturbation uncertainty: when outputs contain an unobserved symmetry, run-to-run variation potentially reflects symmetry rather than error.1
• For point maps, this research derives a closed-form, error-independent variance term that grows with scene extent and potentially overwhelms the desired signal.1
• Simulation reproduces the effect; all 30 real VGGT view-sets tested exhibit its predicted $\|xp\|^2$ signature.1
• ReVar3R robustly registers predictions to a common similarity frame before computing per-point variance, without retraining or modifying the frozen model.1

Sources (1)

• [1]Revar3r: gauge-aware perturbation uncertainty for feed-forward 3d reconstruction
  arXiv (AI, ML, NLP, CV, robotics, multi-agent) · Oct 6, 07:29 AM

  A correctly reconstructed distant point appears uncertain even when a frozen 3D model processes equivalent inputs because its output frame rotates fractionally.

  This exposes a weakness of trainingfree perturbation uncertainty: when outputs contain an unobserved symmetry, run-to-run variation potentially reflects symmetry rather than error.

Extractive summary: sentences quoted from the sources.