AI Paper Atlas

Research map / Detection and segmentation / Tracking

TrackingNet: A Large-Scale Dataset and Benchmark for Object Tracking in the Wild

Matthias Muller, Adel Bibi, Silvio Giancola, Salman Alsubaihi, Bernard Ghanem

ECCV 2018 · Detection and segmentation › Tracking · approach: Memory and correlation-based MOT · 1,237 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
use linear-decay weighted forward/backward average instead of forward pass only (ablation)ECO · OTB100 1-sec sequencesOPE success: 0.805 -> 0.843Improved
use linear-decay weighted average instead of forward pass only (ablation)DCF · OTB100 1-sec sequencesOPE success: 0.771 -> 0.825Improved
use linear-decay weighted average instead of forward pass only (ablation)STAPLE CA · OTB100 1-sec sequencesOPE success: 0.799 -> 0.841Improved
use linear weighting instead of plain 0.5 average of passes (ablation)ECO · OTB100 1-sec sequencesOPE success: 0.824 -> 0.843Improved
use linear weighting instead of exponential decay (ablation)STAPLE CA · OTB100 1-sec sequencesOPE success: 0.835 -> 0.841No clear effect
combine forward and backward pass (average) instead of forward only (ablation)DCF · OTB100 1-sec sequencesOPE success: 0.771 -> 0.799Improved
fine-tune on one of twelve TrackingNet chunks (2511 videos)SiameseFC · OTB100success: 0.569 -> 0.576Improved
fine-tune on one TrackingNet chunkSiameseFC · OTB100precision: 0.765 -> 0.781Improved
fine-tune on one TrackingNet chunkSiameseFC · TrackingNet Testsuccess: 0.571 -> 0.581Improved
fine-tune on one TrackingNet chunkSiameseFC · TrackingNet Testprecision: 0.533 -> 0.543Improved
fine-tune on one TrackingNet chunkSiameseFC · TrackingNet Testnormalized precision: 0.663 -> 0.673Improved
subsample frames by tracker speed (real-time constraint)Struck · TrackingNet Testsuccess: as much as 50%Hurt

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