Stacked Capsule Autoencoders pytorch implementation

Unofficial pytorch implementation of paper:stacked capsule autoencoders. Still at actively developing, welcome to join to discuss! This repository aims to:

• reproducing original paper in pytorch with recent reversion(1.7)
• writing structured and readable code
• to be closed to original implementation as much as possible

There are also some other pytorch implementations such as this, this and this.

[Notice] This repo is still in progress, feel free to make the pytorch version better!

Progress

• CCAE model
• CCAE loss stuff
• CCAE pytorch dataloader
• CCAE training script
• Visualization for CCAE
• eval script for CCAE
• PCAE
• OCAE
• SCAE loss stuff
• SCAE training script

Installation

Based on pytorch 1.7 but sure you can use older version, 1.2~1.7 recommended.

git clone https://github.com/QiangZiBro/stacked_capsule_autoencoders.pytorch
pip install requirements.txt
pip install requirements-dev.txt # for debug or coding quality stuff

CCAE experiment

# in project root directory:
mkdir -p saved/imgs 
python scripts/train_ccae.py

LICENSE

MIT License

Copyright (c) 2020 qiangzibro

Permission is hereby granted, free of charge, to any person obtaining a copy of this software and associated documentation files (the "Software"), to deal in the Software without restriction, including without limitation the rights to use, copy, modify, merge, publish, distribute, sublicense, and/or sell copies of the Software, and to permit persons to whom the Software is furnished to do so, subject to the following conditions:

The above copyright notice and this permission notice shall be included in all copies or substantial portions of the Software.

THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE SOFTWARE.

Reference

• victoresque/pytorch-template for deep learning project template, and this project is built on it.
• stacked_capsule_autoencoders original tensorflow version