Issues
- 1
- 0
- 3
- 9
这个数据为啥这么标注呢
#26 opened by cqray1990 - 1
检测到的横竖线中间被截断,或者整条横竖线不能被完全检测,只能检测到横竖线的部分线段如下图
#28 opened by balicheng - 2
将增量训练完的.h5模型文件转换为.weights文件
#10 opened by aleassx - 0
请问表格外框的训练代码在哪里呢?或者请问训练示例里将图片截取到表格外框那是怎么做到的呢
#30 opened by niconchan - 0
数据标注问题
#29 opened by wuxiaolianggit - 1
- 7
- 0
在tabel_line中,加载完模型后,model.predict(np.array([np.array(inputBlob) / 255.0]))计算时候内存增长特别大
#25 opened by balicheng - 11
关于用自己训练数据,table_ceil.py报错
#4 opened by Liuchen-20191106 - 1
- 1
如何让检测到的表格,按行进行逐个排序,而不是乱序的
#22 opened by kaiven11 - 1
- 18
- 0
训练数据集
#21 opened by garfieldcat1985 - 3
- 24
使用train/train训练时,image模块get_random_data处理后,出现负坐标
#15 opened by deping-1 - 14
Issue with accuracy and loss during training
#17 opened by anishk12345 - 11
权重文件下载链接失效
#7 opened by smileace - 1
各位好:如何在train.py中的main 中添加指定使用GPU的代码,代码默认跑cpu
#19 opened by cdycdycdy - 1
请问标注的是直线,还是一个长条细小的矩形框呢?
#13 opened by cqray1990 - 0
fix_table_box_for_table_line 这个函数的作用怎么理解
#14 opened by pekinghk - 1
unet主干网络sigmoid选用
#12 opened by dcdethan - 0
test内存不释放,如何设置batch_size
#11 opened by dcdethan - 0
训练时横竖线的线宽设置成多少个像素合适?
#9 opened by shihaoyin - 0
Datasource / Paper?
#8 opened by sebastianhorwege - 0
- 0
- 2
- 0