AI Paper Atlas

Research map / Video / Video segmentation and tracking

TrackVerse: A Large-Scale Object-Centric Video Dataset for Image-Level Representation Learning

Yibing Wei, Samuel Church, Victor Suciu, Jinhong Lin, Cheng-En Wu, Pedro Morgado

ICCV 2025 · 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
pretrain on TrackVerse tracks with VA-MoCo instead of ImageNet-1K with MoCo (equal budget)ViT-B/16 MoCo · TrackVerse FSLaccuracy (%): 47.7 -> 71.1Improved
pretrain on TrackVerse tracks instead of ImageNet-1KViT-B/16 MoCo · SSv2 Fttop-1 accuracy (%): 56.8 -> 61.8Improved
use object-centric tracks instead of full scene clips (ablation)VA-MoCo · LVIS-IN NNaccuracy (%): 35.3 -> 50.1Improved
use TrackVerse instead of TAO tracking dataset (ablation)VA-MoCo · LVIS-IN NNaccuracy (%): 14.8 -> 50.1Improved
scale TrackVerse subset from 0.4M to 3.8M tracks (ablation)VA-MoCo · TrackVerse FSLaccuracy (%): 55.3 -> 71.1Improved
add variance-aware predictor conditioning on augmentation parameters (ablation)MoCo on TrackVerse (0.4M-CB1000) · TrackVerse NNaccuracy (%): 33.7 -> 39.9Improved
train longer (100 to 600 epochs) on dynamic tracks versus static frames (ablation)MoCo on TrackVerse (0.4M-CB1000) · TrackVerse NNaccuracy (%): 35.96 -> 39.94 -> 40.43Improved
scale TrackVerse subset from 0.4M to 3.8M tracks (ablation)VA-MoCo · DAVIS-2017J&Fm: 65.4 -> 67.1Improved
scale TrackVerse subset from 2.4M to 3.8M tracks (ablation)VA-MoCo · MIT-States Pair Ft.accuracy (%): 50.0 -> 50.0No clear effect
use ImageNet-1K MoCo versus early small TrackVerse subset (0.4M) on MIT-States Pair Ft.ViT-B/16 MoCo · MIT-States Pair Ft.accuracy (%): 46.8 -> 40.8Hurt

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