MLT
Benchmarks
| Task Name | Dataset Name | SOTA Result | Trend | |
|---|---|---|---|---|
| Scene Text Detection | MLT 2017 (test) | Precision87.3 | 25 | |
| Scene Text Recognition | MLT19 | Average Score90.5 | 21 | |
| Scene Text Recognition | MLT 17 | Average Accuracy94.9 | 21 | |
| End-to-end multilingual recognition | MLT 2019 (test) | F-measure (%)58.2 | 15 | |
| Scene Text Detection | MLT 2017 | Precision85.37 | 14 | |
| Text Detection | MLT 2019 (test) | F-score75.5 | 12 | |
| Scene Text Recognition | MLT 2017 (test) | Acc (%)72.2 | 8 | |
| Joint Text Detection and Script Identification | MLT 2017 (test) | F1 Score74.1 | 8 | |
| Open-set Text Recognition | MLT Japanese 2019 (test) | Character Accuracy (Overall)65.34 | 4 | |
| Text Recognition | MLT 2019 (test) | Accuracy66.7 | 3 | |
| Text Detection | MLT | Recall57.54 | 3 | |
| Object Detection | MLT (test) | Recall58.91 | 3 | |
| Scene Text Recognition | MLT 2017 (val) | Character Accuracy64.1 | 3 |
Showing 13 of 13 rows