Method: A two-stage text detector based on cascade rcnn(using total 10000 images of mlt19) - Samples list - Task 1 - Text Localization - ICDAR 2019 Robust Reading Challenge on Multi-lingual scene text detection and recognition - Robust Reading Competition

Page of 488

next >

View

Thumbnails
Table

Sample	Results
	IoU - Global
	Average Precision	Recall	Precision	Hmean
1	29.17%	50.00%	66.67%	57.14%
2	63.23%	66.67%	85.71%	75.00%
3	50.00%	50.00%	75.00%	60.00%
4	100.00%	100.00%	75.00%	85.71%
5	83.33%	83.33%	100.00%	90.91%
6	22.22%	22.22%	66.67%	33.33%
7	29.50%	58.82%	45.45%	51.28%
8	89.38%	91.67%	91.67%	91.67%
9	63.23%	66.67%	85.71%	75.00%
10	42.43%	50.00%	66.67%	57.14%
11	100.00%	100.00%	100.00%	100.00%
12	100.00%	100.00%	100.00%	100.00%
13	88.89%	88.89%	100.00%	94.12%
14	85.71%	85.71%	100.00%	92.31%
15	100.00%	100.00%	100.00%	100.00%
16	65.98%	70.83%	82.93%	76.40%
17	50.00%	50.00%	100.00%	66.67%
18	100.00%	100.00%	100.00%	100.00%
19	87.50%	87.50%	70.00%	77.78%
20	100.00%	100.00%	100.00%	100.00%

Download full table (.csv)

Results