AI Paper Atlas

Research map / Video / Video generation

AnchorWeave: World-Consistent Video Generation with Retrieved Local Spatial Memories

Zun Wang, Han Lin, Jaehong Yoon, Jaemin Cho, Yue Zhang, Mohit Bansal

ECCV 2026 · Video › Video generation · approach: Trajectory-guided motion control · 14 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
condition on local per-frame point clouds instead of a fused global point cloud (ablation)AnchorWeave (CogVideoX-5B) · RE10K+DL3DV partial-revisitPSNR: 16.31 -> 20.96Improved
retrieve more memories per chunk (K=1 to K=4) (ablation)AnchorWeave (CogVideoX-5B) · RE10K+DL3DV partial-revisitPSNR: 19.01 -> 20.01 -> 20.96Improved
use multiple anchors (K=4) instead of single anchor (ablation)AnchorWeave (Wan2.2-5B) · RE10K+DL3DV partial-revisitSSIM: 0.6180 -> 0.6739Improved
add pose-conditioned fusion instead of simple averaging of anchors (ablation)multi-anchor controller · RE10K+DL3DV partial-revisitvisual artifacts (qualitative)Improved
use joint attention across anchors instead of separate attention (ablation)multi-anchor controller · RE10K+DL3DV partial-revisitvisual quality (qualitative)Improved
replace CogVideoX-5B backbone with Wan2.2-5BAnchorWeave · RE10K+DL3DV partial-revisitVBench quality total: 80.30 -> 80.98Improved
use AnchorWeave single-anchor variant vs single-anchor baselinesAnchorWeave (K=1, CogVideoX) · RE10K+DL3DV partial-revisitPSNR: 19.01 vs 18.23 (Gen3C)Improved
use AnchorWeave (CogVideoX) versus SEVA on reconstruction fidelityAnchorWeave · RE10K+DL3DV partial-revisitPSNR: 20.96 vs 21.13No clear effect
use AnchorWeave versus SEVA on temporal qualityAnchorWeave · RE10K+DL3DV partial-revisitTemporal Flicker: 91.00 -> 93.90Improved
add target camera pose control alongside anchors (no retrieval, first frame only)AnchorWeave · Camera control (300 videos)Translation error: 1.96 -> 1.72 (vs SEVA)Improved
increase K from 1 to 4 (ablation)AnchorWeave (Wan2.2-5B) · EfficiencyGeneration time / GPU memory: 221 -> 275 s; 34 -> 39 GBHurt
fuse more frames into one global point cloud (ablation)global point cloud rendering · Rendering diagnosticPSNR (dB): 44.93 -> 16.15Hurt

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