AI Paper Atlas

Research map / 3D vision / Depth and stereo

$\pi^3$: Permutation-Equivariant Visual Geometry Learning

Yifan Wang, Jianjun Zhou, Haoyi Zhu, Wenzheng Chang, Yang Zhou, Zizun Li, Junyi Chen, Jiangmiao Pang, Chunhua Shen, Tong He

ICLR 2026 · 3D vision › Depth and stereo · approach: Feed-forward 3D reconstruction

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 affine-invariant camera pose modeling (removes reference view, permutation-equivariant) (ablation)Pi3 ablation · ETH3D point mapAccuracy mean (lower better): 0.197 -> 0.131Improved
add scale-invariant local pointmaps (ablation)Pi3 ablation · ETH3D point mapAccuracy mean: 0.229 -> 0.197Improved
add scale-invariant local pointmaps (ablation)Pi3 ablation · 7-Scenes / NRGBD indoorAccuracy: 7-Scenes 0.020 -> 0.020No clear effect
add affine-invariant camera pose modeling (ablation)Pi3 ablation · 7-ScenesAccuracy mean: 0.020 -> 0.019No clear effect
replace reference-view design with permutation-equivariant designVGGT-style baseline · DTU input-order robustnessAcc. std: 0.033 -> 0.003Improved
replace reference-view design with permutation-equivariant designVGGT-style baseline · ETH3D input-order robustnessAcc. std: 0.049 -> 0.000Improved
train from scratch with only core objectives (no global proxy)Pi3 · ETH3D point mapAccuracy: 0.563 (VGGT) vs 0.618Hurt
add auxiliary global pointmap proxy head during from-scratch trainingPi3 · ETH3D point mapAccuracy: 0.618 -> 0.418Improved
add auxiliary global proxy headPi3 · 7-ScenesCompletion: 0.068 -> 0.071No clear effect
Pi3 vs VGGT on camera posePi3 · SintelATE: 0.167 -> 0.074Improved
Pi3 vs VGGT on video depthPi3 · SintelAbs Rel: 0.299 -> 0.233Improved
Pi3 vs VGGT on TUM-dynamics posePi3 · TUM-dynamicsATE: 0.012 vs 0.014No clear effect

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