About environments about anomalydetectioncvpr2018

[roggerfq]

I recommend using virtual environment with pyvenv, once you are in the virtual environment follow the following instructions to install qt4:

http://amyboyle.ninja/Python-Qt-and-virtualenv-in-linux

Do not forget to configure the file "$ HOME / .keras / keras.json" and set "theano" as back-end:

{
     "image_data_format": "channels_last",
     "epsilon": 1e-07,
     "floatx": "float32",
     "backend": "theano"
}

Finally, replace lines 134 and 135 of the "Demo_GUI.py" file:

         weights_path = Model_dir + 'weights_L1L2.mat'
         model_path = Model_dir + 'model.json'

Replace with:

         weights_path = 'weights_L1L2.mat'
         model_path = 'model.json'

[WaqasSultani]

Hi I have used the following

Keras version 1.1.0

Theano 1.0.2

Python 3

Ubuntu 16.04

[henniekim]

Thanks for quick reply :) but actually I would like to know about the details (e.g OpenCV version, pyQT version etc...)

[Hgeun]

We have this error.

[indrango]

You can refactor PyQT4 to PyQT5 to running demo script

[WaqasSultani]

Ok. Thanks