AI Paper Atlas

Research map / Video / Video generation

VGGT-World: Transforming VGGT into an Autoregressive Geometry World Model

Xiangyu Sun, Shijie Wang, Fengyi Zhang, Lin Liu, Caiyan Jia, Ziying Song, Zi Helen Huang, Yadan Luo

ECCV 2026 · Video › Video generation · approach: World-model video generation · 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
predict clean latent (z-prediction) instead of velocity in 1024-d VGGT feature space (ablation)VGGT-World flow transformer · VGGT layer-4 latent trainingdenoising quality / SNRImproved
add Stage-2 trajectory-consistent flow forcing (ablation)VGGT-World · Cityscapes mid-termAbsRel: 0.131 -> 0.126Improved
add Stage-2 trajectory-consistent flow forcing (ablation)VGGT-World · Cityscapes mid-termdelta1: 86.8 -> 87.3Improved
add Stage-2 trajectory-consistent flow forcing (ablation)VGGT-World · TartanAir 1-viewChamfer Distance: 3.0818 -> 2.8451Improved
use linear lambda schedule 0 to 1 instead of static lambda=0.1 (ablation)VGGT-World flow forcing · Cityscapes mid-termAbsRel: 0.142 -> 0.126Improved
use larger static mixing ratio (0.1 -> 0.3 -> 0.7) (ablation)VGGT-World flow forcing · Cityscapes mid-termAbsRel: 0.142 -> 0.148 -> 0.165Hurt
geometry latent state (VGGT) instead of pixel-space video generation (Cosmos)world model for depth forecasting · KITTI short-term meanAbsRel: 0.154 -> 0.065Improved
forecast in frozen VGGT feature space instead of DINO feature space (DINO-Foresight)latent world model · KITTI mid-term meanAbsRel: 0.115 -> 0.098Improved
use geometry latent forecasting instead of Gen3R video generation, pose-freepoint cloud forecasting · TartanAir 1-viewChamfer Distance: 5.7478 -> 2.8451Improved
compare against Gen3R given text conditioningVGGT-World (pose-free) · TartanAir 1-viewCompleteness/CD: Comp. 4.5297 vs 2.5367; CD 2.8451 vs 2.7809Mixed
forecast future horizon and jointly decode with the present frame (ablation)VGGT-World · TartanAir short-horizonChamfer DistanceImproved
extend forecast horizon beyond about 7 frames (ablation)VGGT-World joint decoding · TartanAirChamfer DistanceMixed

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