AI Paper Atlas

Research map / 3D vision / Depth and stereo

Geometric Context Transformer for Streaming 3D Reconstruction

Lin-Zhuo Chen, Jian Gao, Shangzhan Zhang, Yihang Chen, Nan Xue, Jianyuan Wang, Christian Rupprecht, Xun Cao, Xing Zhu, Yujun Shen, Yao Yao, YINGHAO XU

ECCV 2026 · 3D vision › Depth and stereo · approach: Feed-forward 3D reconstruction · 53 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 anchor initialization (ablation)GCT baseline · TartanGround valAUC@3: 9.80 to 13.63; ATE 8.59 to 7.88Improved
add relative pose loss (ablation)GCT · TartanGround valATE: 8.25 -> 7.46; RPE-rot 5.35 -> 2.26Improved
add video RoPE on top of context tokens (ablation)GCT · TartanGround valATE: 7.46 -> 5.98; AUC@3 15.75 -> 16.39Improved
add context tokens (trajectory memory) (ablation)GCT · TartanGround valAUC@30: 68.25 -> 69.92Improved
increase window size from 16 to 64 (ablation)GCT · Oxford SpiresATE: 11.2 -> 5.3 (n=8)Improved
increase window size from 64 to 128 (ablation)GCT · Oxford SpiresATE: 5.5 -> 5.6 (n=2); 5.3 -> 5.3 (n=8)No clear effect
increase anchor count n from 2 to 8 (ablation)GCT · Oxford SpiresATE: 5.5 -> 5.3 (k=64)No clear effect
go from sparse (320 frames) to dense (3,840 frames) sequencesGCT vs CUT3R · Oxford SpiresATE: GCT 5.37 -> 5.60; CUT3R 18.16 -> 32.47Improved
paged KV-cache with FlashInfer instead of contiguous PyTorch cacheGCT · 518x378 video inferenceFPS: ~10.5 FPS -> ~20 FPSImproved
replace causal attention with GCAGCT · context size, T=10,000tokens in context: ~5x10^6 -> ~7x10^4Improved

Earlier methods this paper compares with

Named baselines from earlier accepted papers in this paper's results tables: better than 3 of 3 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