AI Paper Atlas

Research map / 3D vision / Depth and stereo

GGPT: Geometry-Grounded Point Transformer

Yutong Chen, Yiming Wang, Xucong Zhang, Sergey Prokudin, Siyu Tang

CVPR 2026 · 3D vision › Depth and stereo · approach: Foundation models for monocular geometry · 1 citations (OpenAlex)

Read the paperCheck your idea against 137,217 papers

AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.

Measured effects reported in the paper

What was changedWhereMeasuredEffect
refine VGGT dense prediction with GGPTVGGT · ScanNet++ 8 viewsAUC@5/10 cm: 19/32 -> 45/60Improved
refine VGGT with GGPTVGGT · ETH3D 8 views (cross-domain)AUC@5/10 cm: 23/36 -> 47/61Improved
refine Pi3 with GGPT on its training domainPi3 · ScanNet++ 8 viewsAUC@5/10 cm: 56/71 -> 56/72No clear effect
refine Pi3 with GGPTPi3 · ETH3D 8 viewsAUC@5/10 cm: 25/41 -> 36/53Improved
refine MapAnything with GGPTMapAnything · ETH3D 8 viewsAUC@5/10 cm: 7/15 -> 33/45Improved
refine VGGT with GGPTVGGT · 4D-DRESS (human)AUC@1/5 cm: 10/45 -> 66/77Improved
refine MAtCha with GGPTMAtCha · ScanNet++ 4 viewsAUC@5/10 cm: 12/19 -> 27/37Improved
use our SfM guidance instead of MASt3R-SfM guidance (ablation)VGGT + GGPT · ETH3D 8 viewsAUC@5/10 cm: 36/50 -> 47/61Improved
use our SfM guidance instead of MASt3R-SfM guidance (ablation)OMNI-DC · ETH3D 8 viewsAUC@5/10 cm: 25/44 -> 31/44Mixed
use 3D point transformer GGPT instead of 2D depth completion under same guidance (ablation)geometry-conditioned models · ETH3D 8 viewsAUC@5/10 cm: 32/45 (POW3R) -> 47/61 (GGPT)Improved

Earlier methods this paper compares with

Named baselines from earlier accepted papers in this paper's results tables: better than 1 of 1 on most shared results.

Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.

Closest papers by meaning

Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy