我们自己的对象的张量流对象检测

Question

我正在使用 tensorflow 1.9 进行自定义对象检测，并遵循与 https://tensorflow-object-detection-api-tutorial.readthedocs.io/en/latest/training.html#preparing-workspace 中相同的步骤。

但是在训练模型时，我遇到了错误。

(tensorflow_cpu) C:\Users\Z004032A\Documents\Tensorflow\workspace\training_demo>python train.py --logtostderr --train_dir=training/ --pipeline_config_path=training/ssd_inception_v2_coco.config
WARNING:tensorflow:From C:\Users\Z004032A\anaconda3\envs\tensorflow_cpu\lib\site-packages\tensorflow\python\platform\app.py:125: main (from __main__) is deprecated and will be removed in a future version.
Instructions for updating:
Use object_detection/model_main.py.
Traceback (most recent call last):
  File "C:\Users\Z004032A\anaconda3\envs\tensorflow_cpu\lib\site-packages\google\protobuf\text_format.py", line 1460, in _ConsumeSingleByteString
    result = text_encoding.CUnescape(text[1:-1])
  File "C:\Users\Z004032A\anaconda3\envs\tensorflow_cpu\lib\site-packages\google\protobuf\text_encoding.py", line 115, in CUnescape
    .decode('unicode_escape')
UnicodeDecodeError: 'unicodeescape' codec can't decode bytes in position 2-3: truncated \UXXXXXXXX escape

在处理上述异常的过程中，又发生了一个异常：

Traceback (most recent call last):
  File "train.py", line 184, in <module>
    tf.app.run()
  File "C:\Users\Z004032A\anaconda3\envs\tensorflow_cpu\lib\site-packages\tensorflow\python\platform\app.py", line 125, in run
    _sys.exit(main(argv))
  File "C:\Users\Z004032A\anaconda3\envs\tensorflow_cpu\lib\site-packages\tensorflow\python\util\deprecation.py", line 250, in new_func
    return func(*args, **kwargs)
  File "train.py", line 93, in main
    FLAGS.pipeline_config_path)
  File "C:\Users\Z004032A\Documents\Tensorflow\models\research\object_detection\utils\config_util.py", line 100, in get_configs_from_pipeline_file
    text_format.Merge(proto_str, pipeline_config)
  File "C:\Users\Z004032A\anaconda3\envs\tensorflow_cpu\lib\site-packages\google\protobuf\text_format.py", line 685, in Merge
    allow_unknown_field=allow_unknown_field)
  File "C:\Users\Z004032A\anaconda3\envs\tensorflow_cpu\lib\site-packages\google\protobuf\text_format.py", line 752, in MergeLines
    return parser.MergeLines(lines, message)
  File "C:\Users\Z004032A\anaconda3\envs\tensorflow_cpu\lib\site-packages\google\protobuf\text_format.py", line 777, in MergeLines
    self._ParseOrMerge(lines, message)
  File "C:\Users\Z004032A\anaconda3\envs\tensorflow_cpu\lib\site-packages\google\protobuf\text_format.py", line 799, in _ParseOrMerge
    self._MergeField(tokenizer, message)
  File "C:\Users\Z004032A\anaconda3\envs\tensorflow_cpu\lib\site-packages\google\protobuf\text_format.py", line 924, in _MergeField
    merger(tokenizer, message, field)
  File "C:\Users\Z004032A\anaconda3\envs\tensorflow_cpu\lib\site-packages\google\protobuf\text_format.py", line 998, in _MergeMessageField
    self._MergeField(tokenizer, sub_message)
  File "C:\Users\Z004032A\anaconda3\envs\tensorflow_cpu\lib\site-packages\google\protobuf\text_format.py", line 924, in _MergeField
    merger(tokenizer, message, field)
  File "C:\Users\Z004032A\anaconda3\envs\tensorflow_cpu\lib\site-packages\google\protobuf\text_format.py", line 998, in _MergeMessageField
    self._MergeField(tokenizer, sub_message)
  File "C:\Users\Z004032A\anaconda3\envs\tensorflow_cpu\lib\site-packages\google\protobuf\text_format.py", line 924, in _MergeField
    merger(tokenizer, message, field)
  File "C:\Users\Z004032A\anaconda3\envs\tensorflow_cpu\lib\site-packages\google\protobuf\text_format.py", line 1049, in _MergeScalarField
    value = tokenizer.ConsumeString()
  File "C:\Users\Z004032A\anaconda3\envs\tensorflow_cpu\lib\site-packages\google\protobuf\text_format.py", line 1420, in ConsumeString
    the_bytes = self.ConsumeByteString()
  File "C:\Users\Z004032A\anaconda3\envs\tensorflow_cpu\lib\site-packages\google\protobuf\text_format.py", line 1435, in ConsumeByteString
    the_list = [self._ConsumeSingleByteString()]
  File "C:\Users\Z004032A\anaconda3\envs\tensorflow_cpu\lib\site-packages\google\protobuf\text_format.py", line 1462, in _ConsumeSingleByteString
    raise self.ParseError(str(e))
google.protobuf.text_format.ParseError: 170:17 : '    input_path: "C:\Users\Z004032A\Documents\Tensorflow\workspace\trai': 'unicodeescape' codec can't decode bytes in position 2-3: truncated \UXXXXXXXX escape

谁能帮我解决这个问题，并告诉我哪个 tensorflow 版本最适合自定义对象检测

Answer 1

这可能是由于 Windows 使用 C:\ 作为默认用户目录造成的。当您在 Python 中以字符串形式使用此用户目录时，您可能会遇到 unicode 错误，因为使用了 \U unicode 转义。

尝试复制反斜杠。也就是说，把C:\User\Documents 变成C:\\User\\Documents。

至于哪个 Tensorflow 版本最好，没有“最好”的版本。我建议使用与您正在使用的任何库相同的 TF 版本。我还建议不要在原始 TF 中对此进行编码。相反，请使用现有的库，例如 YOLO。只需谷歌“最佳对象检测库 tensorflow”并选择现有库之一。