https://github.com/neka-nat/vin-keras

This is an implimentation of Value Iteration Networks (NIPS2016 best paper) in keras
https://github.com/neka-nat/vin-keras

deep-learning keras nips-2016 value-iteration value-iteration-network

Last synced: 3 months ago
JSON representation

This is an implimentation of Value Iteration Networks (NIPS2016 best paper) in keras

• Host: GitHub
• URL: https://github.com/neka-nat/vin-keras
• Owner: neka-nat
• Created: 2017-01-08T12:12:55.000Z (almost 8 years ago)
• Default Branch: master
• Last Pushed: 2018-01-06T10:32:53.000Z (almost 7 years ago)
• Last Synced: 2024-05-02T01:53:40.267Z (8 months ago)
• Topics: deep-learning, keras, nips-2016, value-iteration, value-iteration-network
• Language: Python
• Homepage:
• Size: 4.66 MB
• Stars: 18
• Watchers: 4
• Forks: 7
• Open Issues: 0
• Metadata Files:
  • Readme: README.md

Awesome Lists containing this project

README

        
# vin-keras

This is an implimentation of [VIN(Value Iteration Networks)](https://arxiv.org/abs/1602.02867) in keras.

* 16x16 Grid world datasets from [vin(chainer) repository](https://github.com/peisuke/vin)

First, train and save VIN model.

    $ python train.py

Next, predict path on the test data.

    $ python predict.py

An example of the results are shown below.

![result0](images/image0.png)

![result1](images/image1.png)

![result2](images/image2.png)

## Reference

* [VIN(original)](https://github.com/avivt/VIN)

* [tensorflow-value-iteration-networks](https://github.com/TheAbhiKumar/tensorflow-value-iteration-networks)

* [vin(chainer)](https://github.com/peisuke/vin)

## Requirements

* Keras

* python-opencv

* numpy

* joblib