wushilian/CRNN_Attention_OCR_Chinese

Issues

• 这是一次性加载所有训练数据进入内存吗？

  #45 opened 4 years ago by LW-CVer
  0

• 代码出问题，怎么解决

  #44 opened 5 years ago by HIHIHAHEI
  0

• 转pb过程中的问题

  #43 opened 5 years ago by changgongcheng
  0

• 请问训练出来的模型有转过caffemodel么

  #42 opened 5 years ago by changgongcheng
  0

• loss下降问题

  #35 opened 6 years ago by yyfanxing
  5

• Do I need to modify charset in config.py in order to recog Chinese?

  #41 opened 5 years ago by ar8327
  0

• Embedding Dimension

  #39 opened 5 years ago by lmpan
  3

• 您好，请问有没有好的中文数据集

  #18 opened 6 years ago by Banyueqin
  1

• acc一直为0.0

  #40 opened 5 years ago by genghuan2005
  0

• 关于单张图片inference的问题

  #17 opened 7 years ago by yourlovedu
  6

• Where do i put the data to run inference

  #37 opened 5 years ago by jihadbourassi
  1

• attention模型和CTC模型效果有比较过吗

  #36 opened 5 years ago by gzhcv
  3

• 单张图片预测infer.py报错，如何解决？

  #29 opened 5 years ago by MatrixPlayer
  2

• 你好，我训练出来的模型只能识别和训练图片一样长度的文本，是某个参数设置错了吗？

  #38 opened 5 years ago by 584769070
  0

• 你好，请问怎么预测不是10个字的图片啊

  #27 opened 6 years ago by sjf18
  2

• could you show the addrress of model trained for us ,thanks?

  #34 opened 6 years ago by maicaijian
  1

• Is this trainning process right?

  #33 opened 6 years ago by lipond
  0

• 模型

  #32 opened 6 years ago by hsl20130659
  0

• synthetic Chinese string dataset

  #31 opened 6 years ago by cloudfool
  3

• Train.py get killed by ubuntu

  #30 opened 6 years ago by yogixRush
  0

• Error

  #16 opened 7 years ago by olfaa
  4

• How to increase training rate.

  #26 opened 6 years ago by SJHBXShub
  3

• dic.txt

  #28 opened 6 years ago by gydlcc
  1

• loss,train_decode_result,pred_decode_result=build_network(is_training=True),这个train_decode_result完全没用到啊？

  #14 opened 6 years ago by zhangtao22
  3

• 训练结果总是 '的'

  #15 opened 6 years ago by badbubble
  21

• 你好，这个程序可以训练中文图片并测试中文吗？我看到config.py里面把英文字母转换成数字，中文的该怎么做呢？谢谢

  #19 opened 6 years ago by 12357abceg
  3

• 报错 Can not squeeze dim[2]

  #20 opened 6 years ago by huyi1989
  2

• 您好，可以提供下您训练好的模型存档吗？就是saved model

  #21 opened 6 years ago by yogixRush
  3

• 有MultiGPU版本吗

  #23 opened 6 years ago by carolinebear
  0

• batchnormolization 没有更新

  #22 opened 6 years ago by hityzy1122
  1

• 训练结果全部变成<GO><GO><GO>

  #25 opened 6 years ago by l00350898
  1

• 不能训练问题

  #13 opened 6 years ago by ElagJun
  9

• an error

  #24 opened 6 years ago by SJHBXShub
  2

• 转成中文识别的错误

  #6 opened 7 years ago by Banyueqin
  16

• MAX_LEN_WORD issue

  #10 opened 7 years ago by cvtower
  2

• How is the acc?

  #11 opened 7 years ago by hanglu
  2

• 请问你这样把宽度固定为120，效果好嘛？会不会让宽高比变形太大

  #12 opened 7 years ago by zhangtao22
  4

• 请问训练和测试时候，decoer的输入不一样吧？

  #9 opened 7 years ago by ccnankai
  3

• trian.py中的acc输出的值是什么的准确率呢？

  #7 opened 7 years ago by lmolhw5252
  1

• 要改成能够识别中文怎么弄呢

  #5 opened 7 years ago by Banyueqin
  6

• run error

  #1 opened 7 years ago by Banyueqin
  2

• 关于生成.pb文件

  #2 opened 7 years ago by Banyueqin
  2

• 怎么预测

  #3 opened 7 years ago by Banyueqin
  11

• 准确率acc一直是0.0

  #4 opened 7 years ago by lmolhw5252
  1