AI Paper Atlas

Research map / 3D vision / 3D generation

Multi-View Foundation Models

Leo Segre, Or Hirschorn, Shai Avidan

ECCV 2026 · 3D vision › 3D generation · approach: Learning 3D reconstruction from images · 3 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
add LoRA training with correspondence loss (ablation)DINOv2 · ScanNet++ correspondencelocalization error: 0.1029 -> 0.0347Improved
add multi-view adapter (ablation)DINOv2 + LoRA · ScanNet++ correspondencelocalization error: 0.0347 -> 0.0262Improved
add multi-view adapter (ablation)DINOv2 + LoRA · base similaritycosine similarity: 0.9117 -> 0.9360Improved
add Plucker ray embedding (ablation)multi-view adapter · ScanNet++ correspondencelocalization error: 0.0262 -> 0.0247Improved
add regularization to base features (ablation)full method · base similaritycosine similarity: 0.1054 -> 0.9376Improved
add regularization to base features (ablation)full method · ScanNet++ correspondencelocalization error: 0.0204 -> 0.0247Hurt
remove camera poses (pose-free variant) (ablation)Ours · DINOv2 similarity generalizationcosine similarity: 0.8003 -> 0.8099No clear effect
multi-view lifting vs FiT3DDINOv2 · NAVI normalsRMSE: 51.00 -> 32.17Improved
multi-view lifting vs SnDDINOv2 · ImageNet zero-shot KNNtop-1 accuracy: 56.31% -> 58.02%Improved
lifting method applied to CLIPCLIP · ScanNet++ correspondencelocalization error: 0.1505 -> 0.0301Improved
lifting method applied to SAMSAM · generalization correspondencelocalization error: 0.1833 -> 0.0752Improved
add meaningful context views from 2 to 10Ours · ScanNet++ correspondencelocalization error: 0.033 -> 0.029Improved

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