# Introduction

## Multimodal Keras Wrapper
Wrapper for Keras with support to easy multimodal data and models loading and handling.

You can download and contribute to the code downloading [this repository](https://github.com/MarcBS/multimodal_keras_wrapper).


## Documentation

You can access the library documentation page at [marcbs.github.io/multimodal_keras_wrapper/](http://marcbs.github.io/multimodal_keras_wrapper/)

Some code examples are available in demo.ipynb and test.py. Additionally, in the section Projects you can see some practical examples of projects using this library.


## Dependencies

The following dependencies are required for using this library:

 - [Anaconda](https://www.continuum.io/downloads)
 - Keras - [custom fork](https://github.com/MarcBS/keras) or [original version](https://github.com/fchollet/keras)
 - [cloud](https://pypi.python.org/pypi/cloud/2.8.5) >= 2.8.5
 - [scipy](https://pypi.python.org/pypi/scipy/0.7.0)

Only when using NMS for certain localization utilities:
 - [cython](https://pypi.python.org/pypi/Cython/0.25.2) >= 0.23.4

## Installation

In order to install the library you just have to follow these steps:

1) Clone this repository:
```
git clone https://github.com/MarcBS/multimodal_keras_wrapper.git
```

2) Include the repository path into your PYTHONPATH:
```
export PYTHONPATH=$PYTHONPATH:/path/to/multimodal_keras_wrapper
```

3) If you wish to install the dependencies (it will install our [custom Keras fork](https://github.com/MarcBS/keras)):
```
pip install -r requirements.txt
```

## Projects

You can see more practical examples in projects which use this library:

[VIBIKNet for Visual Question Answering](https://github.com/MarcBS/VIBIKNet)

[ABiViRNet for Video Description](https://github.com/lvapeab/ABiViRNet)

[Sentence-SelectioNN for Domain Adaptation in SMT](https://github.com/lvapeab/sentence-selectioNN)


## Keras

For additional information on the Deep Learning library, visit the official web page www.keras.io or the GitHub repository https://github.com/fchollet/keras.

You can also use our [custom Keras version](https://github.com/MarcBS/keras), which provides several additional layers for Multimodal Learning.