图像预测问题

Question

我使用 CNN 制作了一个模型来预测给定的 X 射线是否属于 Covid 患者。

我面临的问题

1. 有些图像可以正常工作，而有些则引发错误，我无法确定问题所在

这是错误

ValueError: in user code:

/home/khushwant/.local/lib/python3.8/site-packages/tensorflow/python/keras/engine/training.py:1478 predict_function *
return step_function(self, iterator)
/home/khushwant/.local/lib/python3.8/site-packages/tensorflow/python/keras/engine/training.py:1468 step_function **
outputs = model.distribute_strategy.run(run_step, args=(data,))
/home/khushwant/.local/lib/python3.8/site-packages/tensorflow/python/distribute/distribute_lib.py:1259 run
return self._extended.call_for_each_replica(fn, args=args, kwargs=kwargs)
/home/khushwant/.local/lib/python3.8/site-packages/tensorflow/python/distribute/distribute_lib.py:2730 call_for_each_replica
return self._call_for_each_replica(fn, args, kwargs)
/home/khushwant/.local/lib/python3.8/site-packages/tensorflow/python/distribute/distribute_lib.py:3417 _call_for_each_replica
return fn(*args, **kwargs)
/home/khushwant/.local/lib/python3.8/site-packages/tensorflow/python/keras/engine/training.py:1461 run_step **
outputs = model.predict_step(data)
/home/khushwant/.local/lib/python3.8/site-packages/tensorflow/python/keras/engine/training.py:1434 predict_step
return self(x, training=False)
/home/khushwant/.local/lib/python3.8/site-packages/tensorflow/python/keras/engine/base_layer.py:998 __call__
input_spec.assert_input_compatibility(self.input_spec, inputs, self.name)
/home/khushwant/.local/lib/python3.8/site-packages/tensorflow/python/keras/engine/input_spec.py:255 assert_input_compatibility
raise ValueError(

ValueError: Input 0 of layer sequential is incompatible with the layer: expected axis -1 of input shape to have value 3 but received input with shape (None, 64, 64, 1)

如何重新创建问题？

1. 克隆回购Repo link
2. 进入文件夹并安装requirements.txt
3. 运行 cdclass.py
4. 尝试上传我放在试用文件夹中的图片

您会看到 corona 文件夹中的图像可以正常工作，但是一旦我从普通文件夹上传内容，它就会引发错误，如上所述。

PS：-我尝试了模型中的预测（在我创建它时在 google colab 中）相同的图像工作得很好，但在部署中，它会抛出错误）你会发现一个 . Repo 中的 ipynb 文件，如果您愿意，可以随意查看。

Answer 1

您的某些图像似乎不是RGB图像，而是灰度图像

ValueError: Input 0 of layer sequential is incompatible with the layer: expected axis -1 of input shape to have value 3 but received input with shape (None, 64, 64, 1)

错误告诉您，您的模型希望输入具有 3 个颜色通道，但错误图像只有一个（由 (None, 64, 64, 1) 中的 1 给出）。

这是与此处描述的类似问题：is incompatible with the layer: expected axis -1 of input shape to have value 1 but received input with shape [None, 256, 256, 3] 但在这种情况下，模型似乎缺少正确的输入层。