Papers › Convolutional Character Networks

Convolutional Character Networks

17 Oct 2019ICCV 2019 10arXiv:1910.07954archive 2025-07-28

Linjie Xing, Zhi Tian, Weilin Huang, Matthew R. Scott

Recent progress has been made on developing a unified framework for joint text detection and recognition in natural images, but existing joint models were mostly built on two-stage framework by involving ROI pooling, which can degrade the performance on recognition task. In this work, we propose convolutional character networks, referred as CharNet, which is an one-stage model that can process two tasks simultaneously in one pass. CharNet directly outputs bounding boxes of words and characters, with corresponding character labels. We utilize character as basic element, allowing us to overcome the main difficulty of existing approaches that attempted to optimize text detection jointly with a RNN-based recognition branch. In addition, we develop an iterative character detection approach able to transform the ability of character detection learned from synthetic data to real-world images. These technical improvements result in a simple, compact, yet powerful one-stage model that works reliably on multi-orientation and curved text. We evaluate CharNet on three standard benchmarks, where it consistently outperforms the state-of-the-art approaches [25, 24] by a large margin, e.g., with improvements of 65.33%->71.08% (with generic lexicon) on ICDAR 2015, and 54.0%->69.23% on Total-Text, on end-to-end text recognition. Code is available at: https://github.com/MalongTech/research-charnet.

PaperPDFConference PDFCode

In Syntology Open this paper in Syntology's Atlas, the map of the papers in Syntology's graph and their citations.

Code

MalongTech/research-charnet mentioned in paperpytorchNOASSERTION report

Repository list and official/mentioned flags are the archive's, frozen 2025-07-28. Reachability, where shown, is from one Syntology probe window (2026-09-16 to 2026-09-18); repositories not probed show nothing. GitHub stars are not tracked.

Code Syntology ran Syntology

Not run by Syntology. Nothing on this page verifies that the listed code works.

Tasks

Scene Text DetectionText Detection

Results from the paper archive 2025-07-28

TaskDatasetModelMetricValueRank at snapshotLeaderboardReport
Scene Text Detection ICDAR 2015 CharNet H-88 (multi-scale) F-Measure 91.55 #2 of 43 Archive leaderboard report
Scene Text Detection ICDAR 2015 CharNet H-88 (multi-scale) Precision 92.65 #2 of 43 Archive leaderboard report
Scene Text Detection ICDAR 2015 CharNet H-88 (multi-scale) Recall 90.47 #2 of 43 Archive leaderboard report
Scene Text Detection ICDAR 2015 CharNet H-88 (single-scale) F-Measure 90.97 #3 of 43 Archive leaderboard report
Scene Text Detection ICDAR 2015 CharNet H-88 (single-scale) Precision 89.99 #3 of 43 Archive leaderboard report
Scene Text Detection ICDAR 2015 CharNet H-88 (single-scale) Recall 91.98 #3 of 43 Archive leaderboard report
Scene Text Detection ICDAR 2015 CharNet H-50 (multi-scale) F-Measure 90.16 #4 of 43 Archive leaderboard report
Scene Text Detection ICDAR 2015 CharNet H-50 (multi-scale) Precision 90.9 #4 of 43 Archive leaderboard report
Scene Text Detection ICDAR 2015 CharNet H-50 (multi-scale) Recall 89.44 #4 of 43 Archive leaderboard report
Scene Text Detection ICDAR 2015 CharNet H-57 (multi-scale) F-Measure 90.06 #6 of 43 Archive leaderboard report
Scene Text Detection ICDAR 2015 CharNet H-57 (multi-scale) Precision 91.43 #6 of 43 Archive leaderboard report
Scene Text Detection ICDAR 2015 CharNet H-57 (multi-scale) Recall 88.74 #6 of 43 Archive leaderboard report
Scene Text Detection ICDAR 2015 CharNet H-50 (single-scale) F-Measure 89.7 #8 of 43 Archive leaderboard report
Scene Text Detection ICDAR 2015 CharNet H-50 (single-scale) Precision 91.15 #8 of 43 Archive leaderboard report
Scene Text Detection ICDAR 2015 CharNet H-50 (single-scale) Recall 88.3 #8 of 43 Archive leaderboard report
Scene Text Detection ICDAR 2015 CharNet H-57 (single-scale) F-Measure 89.66 #9 of 43 Archive leaderboard report
Scene Text Detection ICDAR 2015 CharNet H-57 (single-scale) Precision 88.88 #9 of 43 Archive leaderboard report
Scene Text Detection ICDAR 2015 CharNet H-57 (single-scale) Recall 90.45 #9 of 43 Archive leaderboard report
Scene Text Detection ICDAR 2017 MLT CharNet H-88 F-Measure 75.77% #5 of 14 Archive leaderboard report
Scene Text Detection ICDAR 2017 MLT CharNet H-88 Precision 81.27 #5 of 14 Archive leaderboard report
Scene Text Detection ICDAR 2017 MLT CharNet H-88 Recall 70.97 #5 of 14 Archive leaderboard report
Scene Text Detection ICDAR 2017 MLT CharNet R-50 F-Measure 73.42% #11 of 14 Archive leaderboard report
Scene Text Detection ICDAR 2017 MLT CharNet R-50 Precision 77.07 #11 of 14 Archive leaderboard report
Scene Text Detection ICDAR 2017 MLT CharNet R-50 Recall 70.1 #11 of 14 Archive leaderboard report
Scene Text Detection Total-Text CharNet H-88 (multi-scale) F-Measure 86.5% #7 of 27 Archive leaderboard report
Scene Text Detection Total-Text CharNet H-88 (multi-scale) Precision 88 #7 of 27 Archive leaderboard report
Scene Text Detection Total-Text CharNet H-88 (multi-scale) Recall 85 #7 of 27 Archive leaderboard report
Scene Text Detection Total-Text CharNet H-88 F-Measure 85.6% #11 of 27 Archive leaderboard report
Scene Text Detection Total-Text CharNet H-88 Precision 89.9 #11 of 27 Archive leaderboard report
Scene Text Detection Total-Text CharNet H-88 Recall 81.7 #11 of 27 Archive leaderboard report

Ranks are positions in the archive's leaderboards as they stood at the 2025-07-28 snapshot. Results published since then are not among these rows, so a rank here is not a current standing.

Report a problem or propose a change · a person checks every report against the paper or source before anything changes; decisions are listed on /corrections