AI Paper Atlas

Research map / Video / Video-language understanding

A Bilingual, OpenWorld Video Text Dataset and End-to-end Video Text Spotter with Transformer

Weijia Wu, Debing Zhang, Yuanqiang Cai, Sibo Wang, Jiahong Li, Zhuang Li, Yejun Tang, Hong Zhou

NeurIPS 2021 · Video › Video-language understanding · approach: Video-language pretraining · 11 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 TransVTSpotter instead of the strongest listed baseline detectorBOVText text detection · BOVTextF-score: 80.8 -> 81.7Improved
Use TransVTSpotter instead of DB + CRNNBOVText text tracking · BOVTextID_F1: 48.3 -> 64.7Improved
Use TransVTSpotter instead of DB + RAREBOVText end-to-end spotting · BOVTextID_F1: 34.2 -> 40.8Improved
Train TransVTSpotter on BOVText rather than LSVTDTransVTSpotter · BOVText NewsReport scenarioID_F1: 42.1 -> 70.4Improved
Add angle prediction (ablation)TransVTSpotter · ICDAR2015(video) trackingID_F1: 36.1 -> 57.3Improved
Use cosine loss rather than L1 loss for angle prediction (ablation)TransVTSpotter · ICDAR2015 text detectionF-score: 83.2 -> 84.2Improved
Add tracking query to the object query (ablation)TransVTSpotter · ICDAR2015(video) trackingID_F1: 54.2 -> 57.3Improved
Increase input short side from 480 to 800 (ablation)TransVTSpotter · ICDAR2015(video) trackingID_F1: 52.6 -> 57.3Improved
Use TransVTSpotter instead of FreeICDAR2015(video) tracking · ICDAR2015(video)MOTA: 43.2 -> 44.1Improved
Use TransVTSpotter instead of FreeICDAR2015(video) spotting · ICDAR2015(video)ID_F1: 61.8 -> 61.5Hurt
Use TransVTSpotter rather than Wei et al.Video text tracking · Minetto(video)MOTA: 83.5 -> 84.1Improved

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