https://github.com/southworks/CapsNet-CNTK

A CNTK implementation of CapsNet based on Geoffrey Hinton's paper Dynamic Routing Between Capsules
https://github.com/southworks/CapsNet-CNTK

capsnet capsule cntk deep-learning mnist

Last synced: 4 days ago
JSON representation

A CNTK implementation of CapsNet based on Geoffrey Hinton's paper Dynamic Routing Between Capsules

• Host: GitHub
• URL: https://github.com/southworks/CapsNet-CNTK
• Owner: southworks
• Created: 2018-01-23T15:43:31.000Z (almost 7 years ago)
• Default Branch: master
• Last Pushed: 2018-03-15T10:10:13.000Z (over 6 years ago)
• Last Synced: 2024-05-21T12:51:48.609Z (6 months ago)
• Topics: capsnet, capsule, cntk, deep-learning, mnist
• Language: Python
• Homepage:
• Size: 1.8 MB
• Stars: 6
• Watchers: 7
• Forks: 5
• Open Issues: 2
• Metadata Files:
  • Readme: README.md

Awesome Lists containing this project

• awesome-capsule-networks - Southworkscom/CapsNet-CNTK

README

        
# CapsNet-CNTK

A CNTK implementation of CapsNet based on the paper titled '[Dynamic Routing Between Capsules](https://arxiv.org/abs/1710.09829)' by Sara Sabour, Nicholas Frost and Geoffrey Hinton.

## Requeriments

- [Python >=3.5](https://www.python.org/)

- [CNTK 2.4](https://docs.microsoft.com/en-us/cognitive-toolkit/Setup-Windows-Python?tabs=cntkpy24)

- Tensorboard (optional)

## Architecture



## Training

```

git clone https://github.com/southworkscom/CapsNet-CNTK.git

cd CapsNet-CNTK

python get_data.py

python main.py

```

### Tensorboard

Visualize training progress with tensorboard:

```

tensorboard --logdir tensorboard

```

Then, Open your favorite browser and navigate to http://localhost:6006 (assuming you are running the training at localhost)

## Results

### Test Error

   Method     |   Routing   |   Reconstruction  |  MNIST (%)  |  *Paper*

   :---------|:------:|:---:|:----:|:----:

   Baseline |  -- | -- | --             | *0.39*

   CapsNet  |  3 | no | 0.55 | 0.35 (0.036)

   CapsNet  |  3 | yes| WIP | 0.25 (0.005)

### Reconstruction layer



### Dimension perturbations

||||

||||

## TODO

- Improve performance on CPU

- Complete benchmarks

- Capsule visualization

- Improve the API

- Try other datasets