tensorflow.python.framework.errors_impl.InvalidArgumentError: No such callable

[yinzixie]

Keras 2.3.1
tensorflow 1.13.1
opencv-python 3.4.4.19

When I tried to predict the gesture this error happened:

The issues is caused by k.function() due to the version of keras (I guess):

Unfortunately, I am really new to AI, can you help me to fix it? Thank you very much.

[asingh33]

Let me take a look.

[asingh33]

@yinzixie
It's certainly due to newer Keras version. In order to repro the issue, I had to update the keras version.
I couldn't find any good solution with the latest Keras and Tensorflow version. But even with the latest versions of Tensorflow and Keras, I had to do other changes in driver as some of the function names have been changed.
I would suggest to use the older Keras version based on the requirements.txt file in the github, for now, if possible.
I may have to do more findings to make the code fully compatible with latest Keras version, which might take some time.

[c7chengxuyuan]

I train this work,over 500 epoch,but why I don't get good prediction?




I compared the pictures of the training set,whether it was because my prediction pictures did not have details such as palm print.But I'm using your image processing method.
so what is the reason you don't just use local binarization?like
res = cv2.adaptiveThreshold(blur,255,cv2.ADAPTIVE_THRESH_GAUSSIAN_C,cv2.THRESH_BINARY,11,2)

[asingh33]

@c7chengxuyuan Sorry I was away and couldn't back in time.
Were you able to make any progress regarding your issue ?

[sonalgan]

Hi! Is this a balanced dataset we are using?

[FaysalMehrab]

IndexError: list index out of range
When I'm using the option " 1- Use pretrained model for gesture recognition & layer visualization"

Then it isk me "Which weight file to load (enter the INDEX of it, which starts from 0):"

and then this below error shows up.