MaybeShewill-CV/CRNN_Tensorflow

Issues

• windows生成tfrecords多线程报错 freeze_support()

  #335 opened 5 years ago
  4

• Errors may have originated from an input operation.

  #334 opened 5 years ago
  7

• CKPT files

  #333 opened 5 years ago
  2

• 单通道训练的问题

  #332 opened 5 years ago
  1

• when i run python tools/test_shadownet.py? It stop there no bug no out?why?

  #331 opened 5 years ago
  3

• 怎么提高预测速度

  #330 opened 5 years ago
  2

• get different result

  #329 opened 5 years ago
  7

• How to use tfserver？

  #328 opened 5 years ago
  1

• How about add the code of ResNet and Attention module?

  #327 opened 5 years ago
  1

• @LXLun cv2.resize(image, dsize=tuple(CFG.ARCH.INPUT_SIZE), interpolation=cv2.IMREAD_GRAYSCALE) 建议你多看点基础只是 IMREAD_GRAYSCALE不是插值方法

  #326 opened 5 years ago
  4

• Do you compare different optimizers?

  #325 opened 5 years ago
  1

• low GPU memory-usuage

  #324 opened 5 years ago
  1

• 我在改成单通道训练时发生了以下错误

  #323 opened 5 years ago
  2

• How can I train shadownet without char_dict.json, ord_dict.json, lexicon.txt?

  #322 opened 5 years ago
  1

• 使用预训练模型运行失败

  #321 opened 5 years ago
  2

• python write_tfrecords.py 出错

  #320 opened 5 years ago
  2

• 中文数据集的词库

  #319 opened 5 years ago
  13

• 关于W已经存在的问题

  #318 opened 5 years ago
  0

• 中文数据集Synthetic Chinese String Dataset

  #317 opened 5 years ago
  3

• 请问下有没有改成批量预测的思路

  #316 opened 5 years ago
  19

• 评估时 char_dict_path 与 ord_map_dict_path参数的问题

  #315 opened 5 years ago
  7

• 关于模型finetune和字典的问题

  #314 opened 5 years ago
  3

• train doubts

  #313 opened 5 years ago
  8

• 中文训练出来的模型

  #312 opened 5 years ago
  8

• ValueError: The passed save_path is not a valid checkpoint: model/crnn_syn90k_kr92000/shadownet.ckpt

  #311 opened 5 years ago
  11

• 请问不定长文本在测试过程中如何调整送入网络的参数

  #310 opened 5 years ago
  7

• how to adjust the SEQ_LENGTH ?

  #309 opened 5 years ago
  19

• 使用数据训练后只识别出了文本第一个字

  #308 opened 5 years ago
  18

• CNN特征提取的网络架构好像和论文里面的网络架构不太一样，请问会有什么影响吗？

  #307 opened 5 years ago
  1

• 请问我用自己的数据集训练重新训练了模型，为什么就是识别不出来呢，每次results后面都是空

  #306 opened 5 years ago
  2

• 请问我用自己的数据集识别数字和字母时，训练完成之后想要测试一下，但是weight_path在哪里能找到啊

  #305 opened 5 years ago
  1

• tensorflow.python.framework.errors_impl.OutOfRangeError: End of sequence [[{{node train_IteratorGetNext}} = IteratorGetNext[output_shapes=[[32,32,100,3], <unknown>, [32]], output_types=[DT_FLOAT, DT_VARIANT, DT_STRING], _device="/job:localhost/replica:0/task:0/device:CPU:0"](OneShotIterator)]]

  #304 opened 5 years ago
  20

• Shuffle buffer filled and Segmentation fault (core dumped)

  #303 opened 5 years ago
  8

• 我用自己的数据做了一个300张图的数据集，为什么训练的时候出现这样的错误啊，请问我该怎么改

  #302 opened 5 years ago
  4

• 在新增字体类别的情况下，如何支持继续训练？

  #301 opened 5 years ago
  7

• 请问一下如何提高中文模型的识别精度？

  #300 opened 5 years ago
  7

• 测试中文报错

  #299 opened 5 years ago
  2

• 如果我需要训练车牌这样的，需要怎么做了

  #298 opened 5 years ago
  3

• 有没有人用这个模型训练了自己的数据集啊，比如说识别数字字符，能告诉我怎么把自己的数据集做成类似synth90的格式嘛，我该怎么做

  #297 opened 5 years ago
  2

• Training issue

  #296 opened 5 years ago
  7

• AttributeError: module 'tensorflow' has no attribute 'data' & TypeError: Failed to convert object of type <class 'list'> to Tensor. Contents: [Dimension(None), 32, 100, 3]. Consider casting elements to a supported type.

  #295 opened 5 years ago
  9

• cuda 8.0

  #294 opened 5 years ago
  0

• weights on Synth90k dataset

  #293 opened 6 years ago
  12

• 使用其他数据集训练，训练loss正常，但所有预测结果相同

  #292 opened 6 years ago
  14

• 'The origin synth90k dataset_dir'请问这个路径是什么，是synth90 数据集下载之后直接解压出来的mnt文件嘛

  #291 opened 6 years ago
  1

• 怎么将整个synth90k数据集 并将所有文件提取到根目录中

  #290 opened 6 years ago
  2

• 问一个比较简单的问题，下面的路径我该怎么改

  #289 opened 6 years ago
  1

• 请问为什么我训练出来的模型比自带的模型大

  #288 opened 6 years ago
  26

• python write_tfrecords.py 出错

  #286 opened 6 years ago
  4

• create own dataset

  #285 opened 6 years ago
  16