https://github.com/JunYeopLee/capsule-networks

A Tensorflow(v1.4) implementation of Capsule Networks (Dynamic Routing Between Capsules , https://arxiv.org/abs/1710.09829)
https://github.com/JunYeopLee/capsule-networks

Last synced: about 2 months ago
JSON representation

A Tensorflow(v1.4) implementation of Capsule Networks (Dynamic Routing Between Capsules , https://arxiv.org/abs/1710.09829)

• Host: GitHub
• URL: https://github.com/JunYeopLee/capsule-networks
• Owner: JunYeopLee
• License: apache-2.0
• Created: 2018-01-08T06:00:34.000Z (almost 7 years ago)
• Default Branch: master
• Last Pushed: 2018-05-20T04:14:00.000Z (over 6 years ago)
• Last Synced: 2024-08-01T08:11:41.532Z (4 months ago)
• Language: Python
• Homepage:
• Size: 11.4 MB
• Stars: 200
• Watchers: 21
• Forks: 34
• Open Issues: 0
• Metadata Files:
  • Readme: README.md
  • License: LICENSE

Awesome Lists containing this project

• awesome-CapsNet - JunYeopLee/capsule-networks
• awesome-capsule-networks - JunYeopLee/capsule-networks

README

        
# A Tensorflow implementation of Capsule networks

## Requirements

  * tensorflow >= 1.4 (Probably 1.3 should work, too, though I didn't test it)

  * numpy

  * pillow

  * scipy

## Capsule network

I tried to implement the idea in [Dynamic Routing Between Capsules](https://arxiv.org/abs/1710.09829)



## File description

  * `config.py` includes all hyper parameters that are needed.

  * `utils.py` contains functions regarding loading and saving.

  * `model.py` has all building blocks for capsNet and whole model implementation.

  * `train.py` is for training.

  * `eval.py` is for evaluation.

  

## Usage

### Training

```

$ python train.py

```

### Evaluation

```

$ python eval.py

```

## Training

  

## Result

### Classification

| Epoch     | 10     |

|-----------|--------|

| Test Acc. | 99.278 |

### Reconstruction