AI Paper Atlas

Research map / Video / Video segmentation and tracking

SceneScribe-1M: A Large-Scale Video Dataset with Comprehensive Geometric and Semantic Annotations

Yunnan Wang, Kecheng Zheng, Jianyuan Wang, Minghao Chen, David Novotny, Christian Rupprecht, Yinghao Xu, Xing Zhu, Wenjun Zeng, Xin Jin, Yujun Shen

CVPR 2026 · Video › Video segmentation and tracking · approach: Layered video decomposition and dynamics · 0 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 SceneScribe-1M to TartanAir trainingMoGe · Monocular depth average (scale-invariant)Rel: 6.17 -> 6.14Improved
add SceneScribe-1M to TartanAir trainingMoGe · Monocular depth average (affine-invariant depth)Rel: 4.72 -> 4.68Improved
add SceneScribe-1M to TartanAir trainingMoGe · Monocular depth average (affine-invariant disparity)Rel: 5.76 -> 5.68Improved
add SceneScribe-1M to TartanAir trainingMoGe · Sintel (scale-invariant depth)Rel: 19.3 -> 19.6Hurt
add SceneScribe-1M training dataVGGT · CO3Dv2 and ETH3D pose estimationAUC30: 89.5 -> 89.9Improved
add SceneScribe-1M training dataVGGT · CO3Dv2 and ETH3D point mapOverall: 0.677 -> 0.697Hurt
add SceneScribe-1M training dataMonST3R · Sintel pose estimationATE: 0.108 -> 0.099Improved
add SceneScribe-1M training dataMonST3R · Sintel depth estimationdelta1: 58.5 -> 58.1No clear effect
add SceneScribe-1M training dataCoTracker3 · TAP-Vid 2D trackingmean delta_avg^vis: 76.6 -> 77.4Improved
add SceneScribe-1M training dataSpatialTrackerV2 · TAPVid-3Daverage AJ: 23.25 -> 23.5Improved
add SceneScribe-1M training dataAC3D · RealEstate10K camera controlTransErr: 0.374 -> 0.318Improved
add SceneScribe-1M training dataAC3D · RealEstate10K generation qualityFVD: 38.20 -> 35.15Improved

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