Skip to content

This is the implementation of Highway version of Deep convolutional network, based on https://arxiv.org/pdf/1505.00387.pdf.

Notifications You must be signed in to change notification settings

lucko515/deep-convolutional-highway-network

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

1 Commit
 
 
 
 
 
 
 
 

Repository files navigation

Deep Convolutional Highway network - Tensorflow

The implementation of this network is based on the Highway networks paper.

The Highway Network introduces 2 gates in the normal network layer. One gate is called "Transform" gate. The Transform gate is there to control how much information is going to put through from the activation of that layer. Second gate is called the "Carry" gate. It is just 1 - Transform gate. The Carry gate controls how much of UNMODIFIED informations we pass through the network.

By combining these two gates with normal formula for the layer f(xW + b) we get the next formula:

Note: The tricky part with DC Highway networks are dimensions. Just be conscious about the highway layer tensor dimensions.

Install

    Python version

         - Python version used in this project: 3.5+

    Dependencies

Code

The code for this project is inside highway_dc_network.ipynb.

How to run

To run this project execute one of these two lines in your terminal/cmd:

ipython notebook highway_dc_network.ipynb

or

jupyter notebook highway_dc_network.ipynb

Read more about Highway network

Highway networks paper

HIGHWAY AND RESIDUAL NETWORKS LEARN UNROLLED ITERATIVE ESTIMATION

Bonus read: Convolutional Networks

License

MIT License

Copyright (c) 2017 Luka Anicin

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.

Releases

No releases published

Packages

No packages published