This is VGG16 AutoEncoder developed by Keras and Tensorflow backend.

Requirement

• python3.6
• keras
• Tensorflow
• cv2
• h5

How to use

Under project directory:

• data_source:

  Put training data and testing data there(separated by sub-directory).

• model_and_weight:

  Basically, vgg16_weights_notop.h5 is necessary.

  And generated models and weight will be stored there.

• ModelTrainer.py:

  Run this to train model.(trained through data_source\training and evaluate through data_source\testing)

• AutoEncoderEfficiencyEvaluator.py

  Run this to generate images after encoded and decoded.(data within data_source\testing)

  The images would be stored in auto_encoder sub-directory.