AI Paper Atlas

Research map / Vision-language and multimodal / Human and embodied modeling

From Two to One: A New Scene Text Recognizer With Visual Language Modeling Network

Yuxin Wang, Hongtao Xie, Shancheng Fang, Jing Wang, Shenggao Zhu, Yongdong Zhang

ICCV 2021 · Vision-language and multimodal › Human and embodied modeling · approach: Language-grounded scene text and structure · 193 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
add MLM occluded training (ratio 1:1) (ablation)VRM-2L baseline · IC15accuracy: 79.8 -> 81.8Improved
add MLM occluded training (ratio 1:1) (ablation)VRM-2L baseline · SVTPaccuracy: 81.1 -> 83.7Improved
add MLM occluded training (ratio 1:1) (ablation)VRM-2L baseline · CUTE80accuracy: 85.8 -> 88.2Improved
add MLM occluded training (ratio 1:1) (ablation)VRM-2L baseline · IIIT5Kaccuracy: 94.5 -> 95.4Improved
raise occluded ratio from 1:1 to 2:1 (ablation)MLM · SVTaccuracy: 91.0 -> 90.0Hurt
use both WCL branches instead of Mas only (ablation)MLM · SVTaccuracy: 89.8 -> 91.0Improved
use both WCL branches instead of Rem only (ablation)MLM · SVTPaccuracy: 82.2 -> 83.7Improved
replace dropout/cutout with MLM masking (ablation)masking method · 6-benchmark averageaccuracy: 89.0 -> 90.2Improved
random cutout instead of no masking (ablation)baseline · 6-benchmark averageaccuracy: 88.8 -> 89.0No clear effect
use three transformer units instead of two in VRM (ablation)VRM · IC15accuracy: 81.8 -> 83.7Improved
use three transformer units instead of two in VRM (ablation)VRM · CUTE80accuracy: 88.2 -> 88.5Improved
replace separate language model with VisionLAN (ablation)baseline + Transformer LM · OSTaverage accuracy: 58.2 -> 60.3Improved

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