Zhida Huang, Zhuoyao Zhong, Lei Sun, Qiang Huo
In this paper, we present a new Mask R-CNN based text detection approach which can robustly detect multi-oriented and curved text from natural scene images in a unified manner. To enhance the feature representation ability of Mask R-CNN for text detection tasks, we propose to use the Pyramid Attention Network (PAN) as a new backbone network of Mask R-CNN. Experiments demonstrate that PAN can suppress false alarms caused by text-like backgrounds more effectively. Our proposed approach has achieved superior performance on both multi-oriented (ICDAR-2015, ICDAR-2017 MLT) and curved (SCUT-CTW1500) text detection benchmark tasks by only using single-scale and single-model testing.
| Task | Dataset | Metric | Value | Model |
|---|---|---|---|---|
| Scene Text Detection | SCUT-CTW1500 | F-Measure | 85 | PAN |
| Scene Text Detection | SCUT-CTW1500 | FPS | 65.2 | PAN |
| Scene Text Detection | SCUT-CTW1500 | Precision | 86.8 | PAN |
| Scene Text Detection | SCUT-CTW1500 | Recall | 83.2 | PAN |
| Scene Text Detection | ICDAR 2017 MLT | Precision | 80 | PAN |
| Scene Text Detection | ICDAR 2017 MLT | Recall | 69.8 | PAN |
| Scene Text Detection | ICDAR 2015 | F-Measure | 85.9 | PAN |
| Scene Text Detection | ICDAR 2015 | Precision | 90.8 | PAN |
| Scene Text Detection | ICDAR 2015 | Recall | 81.5 | PAN |