AI Paper Atlas

Research map / 3D vision / 3D generation

STream3R: Scalable Sequential 3D Reconstruction with Causal Transformer

Yushi LAN, Yihang Luo, Fangzhou Hong, Shangchen Zhou, Honghua Chen, Zhaoyang Lyu, Bo Dai, Shuai Yang, Chen Change Loy, Xingang Pan

ICLR 2026 · 3D vision › 3D generation · approach: Learning 3D reconstruction from images

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 RNN state memory (CUT3R) with causal decoder-only attention over cached features (ablation)STream3R-alpha · Sintel video depthAbs Rel: 0.598 -> 0.535Improved
replace RNN state memory with causal decoder-only attention (ablation)STream3R-alpha · KITTI video depthAbs Rel: 0.157 -> 0.141Improved
replace RNN state memory with causal decoder-only attention (ablation)STream3R-alpha · 7-Scenes 3D reconstructionAcc mean: 0.480 -> 0.328Improved
replace RNN state memory with causal decoder-only attention (ablation)STream3R-alpha · training throughputtraining steps in same time: 60% more training stepsImproved
use sliding window attention (window 5) with constant cache (ablation)STream3R-beta · KITTI video depthFPS: 12.95 -> 32.93Improved
use sliding window attention (window 5) with constant cache (ablation)STream3R-beta · Bonn video depthAbs Rel: 0.069 -> 0.064Improved
use sliding window attention (window 5) with constant cache (ablation)STream3R-beta · Sintel video depthAbs Rel: 0.264 -> 0.279Hurt
replace causal attention with full attention at inference (ablation)STream3R-beta · 7-Scenes 3D reconstructionComp mean: 0.110 -> 0.075Improved
initialize from VGG-T instead of DUSt3R weightsSTream3R · Sintel single-frame depthAbs Rel: 0.350 -> 0.228Improved
remove single-view datasets from training (ablation)STream3R-beta · training dataperformanceNo 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