Question

我正在尝试获取tensorflow初始v3的pool_3层输出。我的输入是形状的ndarray（64,64,3），但我得到以下错误

with tf.Session() as sess:
    pool_3_tensor = sess.graph.get_tensor_by_name('pool_3:0')

    feat1 = sess.run(pool_3_tensor,{'DecodeJpeg/contents:0': image})  
    feat1 = np.squeeze(feat1)

ValueError                                Traceback (most recent call last)
<ipython-input-26-fb1865f7fbee> in <module>()
      3     pool_3_tensor = sess.graph.get_tensor_by_name('pool_3:0')
      4 
----> 5     feat1 = sess.run(pool_3_tensor,{'DecodeJpeg/contents:0': image})
      6     feat1 = np.squeeze(feat1)

/N/u/mbirla/anaconda3/envs/tensorflow/lib/python3.5/site-packages/tensorflow/python/client/session.py in run(self, fetches, feed_dict, options, run_metadata)
    765     try:
    766       result = self._run(None, fetches, feed_dict, options_ptr,
--> 767                          run_metadata_ptr)
    768       if run_metadata:
    769         proto_data = 
    tf_session.TF_GetBuffer(run_metadata_ptr)

/N/u/mbirla/anaconda3/envs/tensorflow/lib/python3.5/site-packages/tensorflow/python/client/session.py in _run(self, handle, fetches, feed_dict, options, run_metadata)
    942                 'Cannot feed value of shape %r for Tensor %r, '
    943                 'which has shape %r'
--> 944                 % (np_val.shape, subfeed_t.name, str(subfeed_t.get_shape())))
    945           if not self.graph.is_feedable(subfeed_t):
    946             raise ValueError('Tensor %s may not be fed.' % subfeed_t)

ValueError: Cannot feed value of shape (64, 64, 3) for Tensor 'DecodeJpeg/contents:0', which has shape '()

-------更新----------

将其转换为字符串后。我收到无效的JPEG数据，大小12288。有关详细错误：https://gist.github.com/mridulbirla/0d710d7ccd7b22c8f87989c37837e10e

Answer 1

更改

feat1 = sess.run(pool_3_tensor,{'DecodeJpeg/contents:0': image})

到

feat1 = sess.run(pool_3_tensor,{'DecodeJpeg/contents:0': image.tostring()})

并试一试

'DecodeJpeg/contents:0'是一个标量字符串张量，您可以从中解码三个暗淡的图像张量

当我用inception-v3转移学习时，我没有将图像读入np.array类型，而是将图像内容读取为带有

的字符串

with tf.gfile.FastGFile("you_image_file_name") as f:
     content = f.read()

然后将content提供给'DecodeJpeg/contents:0'

Tensorflow Inception V3在pool_3图层输出时出错

1 个答案: