method: CNN based method2019-09-10
Authors: Geonho Hwang
Affiliation: NCIA, Seoul National University
Description: This method is based on E2E-MLT-an Unconstrained End-to-End Method for Multi-Language Scene Text.
We use ICDAR 2017,2019 training, validation and synthetic data.
We ensemble 3 results from different networks.
Confusion Matrix
| Detection | ||||||||||
|---|---|---|---|---|---|---|---|---|---|---|
| Arabic | Latin | Chinese | Japanese | Korean | Bangla | Symbols | Mixed | None | ||
| GT | Arabic | 4777 | 319 | 2 | 16 | 21 | 2 | 5 | 0 | 0 |
| Latin | 146 | 59466 | 137 | 358 | 311 | 37 | 82 | 0 | 0 | |
| Chinese | 7 | 608 | 3065 | 944 | 115 | 7 | 4 | 0 | 0 | |
| Japanese | 43 | 2751 | 973 | 4055 | 314 | 10 | 11 | 0 | 0 | |
| Korean | 19 | 1951 | 125 | 234 | 10640 | 20 | 3 | 0 | 0 | |
| Bangla | 7 | 186 | 7 | 18 | 24 | 2303 | 0 | 0 | 0 | |
| Symbols | 45 | 1443 | 1 | 29 | 18 | 7 | 1953 | 0 | 0 | |
| Mixed | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | |
| None | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | |