AI Paper Atlas

Research map / Detection and segmentation / CNN features

Scene Text Detection and Recognition: The Deep Learning Era

Shangbang Long, Xin James He, Cong Yao

IJCV 2020 · Detection and segmentation › CNN features · approach: Arbitrary-shape scene text detection · 574 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
newer detection methods over earliest listedscene text detectors · ICDAR 2015F1: 54 -> 89.70Improved
newer detection methods over earliest listedscene text detectors · ICDAR 2013F1: 83 -> 95.2Improved
newer detection methods over earliest listedscene text detectors · MSRA-TD500F1: 66 -> 85.08Improved
newer detection methods over earliest listedscene text detectors · CTW1500F1: 73.4 -> 84.45Improved
ResNet backbone with 90k + SynthText instead of VGG with 90kscene text recognizers · IIIT5k (no lexicon)accuracy: 81.2 -> 93.4Improved
faster methods trade off accuracydetectors · ICDAR 2015F1 vs FPS: 13.2 FPS at 78.20 F1; 10.0 FPS at 87.6 F1Mixed
newer end-to-end methods over earliestend-to-end spotters · ICDAR 2015 (strong lexicon)end-to-end F-score: 81.09 -> 83.38Improved
curved text in training synthetic data (Curved ST)scene text recognizers · Total-TextaccuracyImproved

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