This is an implementation of Autoencoder and image comparison on Python 3, Keras, and Tensorflow. It's based on traditional Autoencoder and Relative entropy. Licensed under GPL-3.0 License (see LICENSE for details)
The repository includes:
- Source and Training code of Convolutional Autoencoder based on traditional Autoencoder.
- Pre-trained weights for CAE.
- Jupyter notebook to specify every step.
- Extract.ipynb Is the easiest way to get started. It shows detailed preprocessing and training steps.
- run.py includes source and training code of Convolutional Autoencoder based on dataset.
- image_compare.py includes code to compare two images.
- datasets includes two sets of data I provide.
- results/model includes two pre-trained weights.
- os
- numpy
- matplotlib
- tensorflow>=1.3.0
- keras>=2.0.8
- opencv-python
- IPython[all]
Author: NPU-Franklin@Github.com
Time: 2020-2-3 16:25:35