9.27---报错整理

1. 错误一：
   1. 报错：

      RuntimeError: tf.gradients is not supported when eager execution is enabled. Use tf.GradientTape instead.

   2. 报错原因：chow_output类型是tf.tensor，last_conv_layer是keras.layers.convolustional.conv2d.Conv2D 出现类型不兼容，查询到的解决方法是直接替换为使用tensorflow中的keras.
   3. 解决：

      原始代码：
           grads=K.gradients(chow_output,last_conv_layer)[0]  # 松狮类别相对于block5_conv3输出特征图的梯度 
      
      修改后的代码：
          with tf.GradientTape() as gtape:
                  grads = gtape.gradient(chow_output, last_conv_layer.output)

2. 错误二：
   1. 报错：

      module 'cv2.cv2' has no attribute 'xfeatures2d'

3. 错误三：
   1. 报错：

      error: (-215:Assertion failed) npoints >= 4 && npoints == std::max(ipoints.checkVector(2, CV_32F), ipoints.checkVector(2, CV_64F)) in function 'cv::solvePnPRansac'

4. 错误四：
   1. 报错：

      The 'color' trait of a GlyphFactory instance must be a tuple of the form: (0.0 <= a floating point number <= 1.0, 0.0 <= a floating point number <= 1.0, 0.0 <= a floating point number <= 1.0) or None, but a value of array([0.70980392, 0.69019608, 0.67843137]) <class 'numpy.ndarray'> was specified.

   2. 解决：报错信息已很明显地提示，解决：只需要将array中的元素逐个拿出然后用括号括起来即可形成tuple
5. 错误五：
   1. 报错：module 'sklearn.datasets' has no attribute 'samples_generator'
   2. 解决：删除samples_generator
6. 错误六：
   1. 报错：报错：ImportError: No module named ‘keras_contrib‘
   2. 解决：pip install git+https://www.github.com/keras-team/keras-contrib.git