AI Paper Atlas

Research map / Detection and segmentation / Tracking

Fast Spatial Tracking with Visual Geometry Transformer

Chengjie Huang, Guile Wu, Dongfeng Bai, Bingbing Liu

CVPR 2026 · Detection and segmentation › Tracking · approach: Transformer-based video object tracking

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 global embedding branch to the frame-only tracker (ablation)Fast Spatial Tracker · TAPVid-3D minivalADP: 14.5 -> 16.0Improved
add multi-layer cross attention (MLCA) to the frame-only tracker (ablation)Fast Spatial Tracker · TAPVid-3D minivalADP: 14.5 -> 15.0Improved
combine global branch and MLCA (ablation)Fast Spatial Tracker · TAPVid-3D minivalADP: 14.5 -> 16.3Improved
add four real-world 3D datasets (ScanNet, ARKitScenes, nuScenes, Argoverse2) to synthetic training (ablation)Fast Spatial Tracker · TAPVid-3D minivalADP: ADP 16.3 -> 21.9; OA 81.5 -> 76.8Mixed
fine-tune DINOv2 and alternating attention layers of the backbone with reduced learning rate (ablation)Fast Spatial Tracker · TAPVid-3D minivalADP: 21.9 -> 25.7Improved
add DynPose100K++ and Wild-SDG1M pseudo-labeled in-the-wild and AI-generated data (ablation)Fast Spatial Tracker · TAPVid-3D minivalADP: 25.7 -> 27.5Improved
switch depth prior from UniDepthV2 to MegaSaMDELTA · TAPVid-3DADP: 20.6 -> 26.3Improved
switch depth prior from UniDepthV2 to MegaSaMSpaTracker · TAPVid-3DADP: 16.8 -> 20.8Improved
remove depth prior and use geometry transformer features directlyFast Spatial Tracker vs SpaTrackerV2 · TAPVid-3D inferencetime per frame: 309ms -> 28msImproved
use the 3D tracker for 2D point trackingFast Spatial Tracker · TAP-VidAJ: 61.4 vs AllTracker 67.8Hurt

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