AI Paper Atlas

Research map / 3D vision / Depth and stereo

TTT3R: 3D Reconstruction as Test-Time Training

Xingyu Chen, Yue Chen, Yuliang Xiu, Andreas Geiger, Anpei Chen

ICLR 2026 · 3D vision › Depth and stereo · approach: Foundation models for monocular geometry · 102 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
replace CUT3R's constant learning rate with confidence-guided per-token learning rate (TTT3R)CUT3R · TUM-dynamic pose, 1000 framesATE: 0.173 -> 0.106Improved
replace CUT3R's constant learning rate with TTT3R ruleCUT3R · KITTI depth, 500 framesdelta<1.25: 80.2 -> 86.9Improved
use learned gating (ScalarLR/ConditionLR/TokenLR) instead of training-free closed-form rate (ablation)CUT3R · TUM-dynamic pose, 1000 framesATE: 0.173 -> 0.154 (TokenLR best) vs 0.106 TTT3RImproved
learned ScalarLR gating (ablation)CUT3R · TUM-dynamic poseRPE rot: 0.494 -> 0.502No clear effect
finetune with TTT3R update rule on 4-64 views (ablation)TTT3R · TUM-dynamic poseATE: 0.106 -> 0.091Improved
finetune with TTT3R update rule on 4-64 views (ablation)TTT3R · KITTI depthAbs Rel: 0.131 -> 0.133Hurt
add periodic state reset every 100 frames to TTT3R (ablation)TTT3R · TUM-dynamic pose, 1000 framesATE: 0.106 -> 0.093Improved
TTT3R vs reset/EMA/burn-in non-TTT baselines (ablation)CUT3R · TUM-dynamic pose, 1000 framesATE: 0.126 (Reset) -> 0.106Improved
EMA shrinkage to initial state (ablation)CUT3R · TUM-dynamic poseATE: 0.173 -> 0.164No clear effect
vary reset period 50/100/150 (ablation)CUT3R + Reset · TUM-dynamic poseATE: 0.129 / 0.126 / 0.145Mixed
TTT3R on short sequences vs CUT3RCUT3R · ScanNet 90 frames poseATE: 0.099 -> 0.064Improved
TTT3R vs offline VGGTTTT3R · ScanNet 90 frames poseATE: 0.035 (VGGT) vs 0.064Hurt

Later papers that compared with this method

10 later accepted papers report TTT3R as a baseline in their results tables; 10 of them report a better number on most of the results they share with it.

2 more not listed.

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