-
File "D:\ocr_demo\bezier_curve_text_spotting-master\maskrcnn_benchmark\layers\nms.py", line 8, in
nms = amp.float_function(_C.nms)
File "D:\Anaconda\envs\pytorch_gpu\lib\site-packages\yacs\…
-
while train the model at 10000 steps at my own datasets, and turn to eval the model, encountered errors as follow:
AttributeError: 'TextEvaluator' object has no attribute '_text_eval_gt_path'
a…
-
Hi,
I am trying to understand `ABCNet` in detail and want to use it for custom data set. I'm pretty new for scene text recognition site though, and by this paper-work, I get informed some of the high…
innat updated
4 years ago
-
按照文章的步骤来做，先在synth80k上训2个epoch，然后再CTW1500上finetune1200个epoch，结果的准确率和文章一样，但是召回率却比文章低了10个点。求解答。
-
你好，请问该方法在其他包括中字符的数据集表现如何？如icpr2018、ReCTs、LSVT、art、ctw1500数据集。谢谢
xiliu updated
4 years ago
-
what do the labels of total_text and scut label look like?
The total_text‘s label is same to CTW1500?
and the scut using the char labels, whose each box has 8 value?
thanks a lot
-
There are some inconsistency of gt labels from CTW-1500-reg-label.zip and ctw1500.zip. For example, in 1017.txt, there are 21 instances in CTW-1500-reg-label.zip and 20 instances in ctw1500.zip.
-
你好，我想请教下这个代码在scut数据集上的效果怎么样，我用这个代码在icdar15数据集上测试过，效果能达到precision:0.8212，recall:0.7694，hmean:0.7944。
有一点不太明白的是论文中提出psenet可以痛过分割检测出任意形状的文字，但是我在看代码中生成mask的时候，是通过边框的四点坐标将原来的边框整个转换为了mask，是否对于scut数据集有不同的处理…
-
```
---------------------- ---------------------------------------------------------
sys.platform linux
Python 3.7.4 (default, Aug 13 2019, 20:35:49) [GCC 7.3.0]
numpy…
-
Thanks for your contribution.
How to train the model. And how to get the training data,
Thanks