关于U-net的执行顺序问题

[myboyliu]

您好： 在看了您的代码后觉得很有帮助，但是readme写的不是很清楚。 目前不知道执行顺序。 2 data augmentation the data and label should be the same change 对于步骤2，先执行data.py，然后执行augdata.py？ 目前出错，错误日志如下： loading data

load train images...

---

load test images...

loading data done /Users/ericliu/cv/detect-cell-edge-use-unet/data.py:231: RuntimeWarning: Mean of empty slice. mean = imgs_test.mean(axis=0) /Users/ericliu/anaconda/lib/python2.7/site-packages/numpy/core/_methods.py:73: RuntimeWarning: invalid value encountered in true_divide ret, rcount, out=ret, casting='unsafe', subok=False) Traceback (most recent call last): File "/Users/ericliu/cv/detect-cell-edge-use-unet/unet_test.py", line 169, in myunet.train() File "/Users/ericliu/cv/detect-cell-edge-use-unet/unet_test.py", line 153, in train model = self.get_unet() File "/Users/ericliu/cv/detect-cell-edge-use-unet/unet_test.py", line 87, in get_unet print("conv1 shape:", conv1.shape) AttributeError: 'Tensor' object has no attribute 'shape'

Process finished with exit code 1

[myboyliu]

看了你的博客，应该是TensorFlow版本的问题。升级为1.2.1后没有那个错误了。

[Manfestain]

得到的predict效果好差，楼主能不能给点建议呀

[Manfestain]

我看了一下，是load_weights没加载上，应该是hdf5库的问题