AI Paper Atlas

Research map / Label-efficient and robust learning / Semi-supervised learning

CoTracker3: Simpler and Better Point Tracking by Pseudo-Labelling Real Videos

Nikita Karaev, Yuri Makarov, Jianyuan Wang, Natalia Neverova, Andrea Vedaldi, Christian Rupprecht

ICCV 2025 · Label-efficient and robust learning › Semi-supervised learning · approach: Pseudo-labeling for specialized tasks · 216 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
fine-tune on 15k real pseudo-labelled videosCoTracker3 online · TAP-Vid meandelta_avg^vis: 75.1 -> 76.1Improved
add cross-track attention (ablation)CoTracker3 online · Dynamic Replicadelta_avg^occ: 35.9 -> 41.0Improved
add cross-track attention (ablation)CoTracker3 online · Dynamic Replicadelta_avg^vis: 71.3 -> 72.9Improved
self-training on own predictions using real videos (ablation)CoTracker3 online · TAP-Vid meanAJ: 62.2 -> 63.5Improved
add more diverse teachers (RT online, RT offline, TAPIR, CoTracker) (ablation)pseudo-labelling · TAP-Vid meandelta_avg: 75.7 -> 76.8Improved
choose a random teacher instead of averaging or median of teachers (ablation)pseudo-labelling · TAP-Vid DAVISdelta_avg: 76.5 -> 77.0Improved
freeze confidence and visibility head during real-video fine-tuning (ablation)CoTracker3 · TAP-Vid averageOA: 86.3 -> 90.2Improved
change point sampling method (Uniform, SuperPoint, DISK, SIFT) (ablation)pseudo-label queries · TAP-Viddelta_avg: DAVIS 76.7 to 77.0No clear effect
scale real video data beyond 30k videosCoTracker3 online and offline, LocoTrack · TAP-Viddelta_avgNo clear effect
repeat the scaling pipeline a second time (ablation)CoTracker3 offline · TAP-Vid averagedelta_avg: 76.8 -> 76.9No clear effect
train longer on 15k videos (60k vs 15k iterations) (ablation)CoTracker3 offline · TAP-Vid averagedelta_avg: 76.8 -> 77.0No clear effect
add real-video fine-tuning to offline model for occluded pointsCoTracker3 online · Dynamic Replicadelta_avg^occ: online 41.0 -> 40.1, offline 41.8 -> 42.3Mixed

Later papers that compared with this method

1 later accepted paper reports CoTracker3 as a baseline in their results tables; 1 of them report a better number on most of the results they share with it.

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