https://github.com/danielhavir/capsule-network

PyTorch implementation of the paper Dynamic Routing Between Capsules by Sara Sabour, Nicholas Frosst and Geoffrey Hinton
https://github.com/danielhavir/capsule-network

Last synced: 8 days ago
JSON representation

PyTorch implementation of the paper Dynamic Routing Between Capsules by Sara Sabour, Nicholas Frosst and Geoffrey Hinton

• Host: GitHub
• URL: https://github.com/danielhavir/capsule-network
• Owner: danielhavir
• License: mit
• Created: 2018-03-02T22:22:35.000Z (over 6 years ago)
• Default Branch: master
• Last Pushed: 2018-05-06T17:22:32.000Z (over 6 years ago)
• Last Synced: 2024-08-01T08:11:41.654Z (3 months ago)
• Language: Python
• Size: 48.8 KB
• Stars: 165
• Watchers: 7
• Forks: 23
• Open Issues: 1
• Metadata Files:
  • Readme: README.md
  • License: LICENSE

Awesome Lists containing this project

• awesome-CapsNet - danielhavir/capsule-network
• awesome-capsule-networks - danielhavir/capsule-network

README

        
# Capsule Network #

[![License](https://img.shields.io/github/license/mashape/apistatus.svg?maxAge=2592000)](LICENSE)

#### PyTorch implementation of the following paper:

* [_Dynamic Routing Between Capsules_](https://arxiv.org/abs/1710.09829) by Sara Sabour, Nicholas Frosst and Geoffrey Hinton

### Official implemenation

* [Official implementation](https://github.com/Sarasra/models/tree/master/research/capsules) (TensorFlow) by Sara Sabour

### Visual represenation

![capsules_visual_representation](https://cdn-images-1.medium.com/max/1600/1*UPaxEd1A3N5ceckB85RIRg.jpeg)

> Image source: _Mike Ross_, [A Visual Representation of Capsule Network Computations](https://medium.com/@mike_ross/a-visual-representation-of-capsule-network-computations-83767d79e737)

### Run the experiment

* For details, run `python main.py --help`

### Example of reconstructed vs. original images

![reconstructed](reconstructed.png)

______

### Requirements:

* PyTorch (http://www.pytorch.org)

* NumPy (http://www.numpy.org/)

* GPU

### Default hyper-parameters (similar to the paper):

* Per-GPU `batch_size` = 128

* Initial `learning_rate` = 0.001

* Exponential `lr_decay` = 0.96

* Number of routing iteration (`num_routing`) = 3

#### Loss function hyper-parameters (see [loss.py](loss.py)):

* Lambda for Margin Loss = 0.5

* Scaling factor for reconstruction loss = 0.0005

### GPU Speed benchmarks:

(with above mentioned hyper-parameters)

* Single GeForce GTX 1080Ti - 35.6s per epoch

* Two GeForce GTX 1080Ti - 35.8s per epoch (twice the batch size -> half the iteration)