You must have the following installed on your machine:
- Node.js & npm (comes with Node.js)
- pip - the python package manager
- openCV - python package for Computer Vision
- avconv - video and audio converter
Once you have cloned the directory, navigate to the project's root directory and run the following command:
$ ./install.sh
This script will install all the required dependencies for the web application and needed python packages.
Note: A computer with a CUDA capable GPU compatible with Theano must be used to run the code and the web app - we used the NVIDIA GEFORCE GTX TITAN X.
In a terminal, navigate to the ADOSAUR/app/Backend directory and run the following command to start the server:
$ ./start.sh
Alternatively, if you have grunt-cli installed globally on your machine, just run:
$ grunt run
Then in a web browser, open the page at http://localhost:8080.
The set of Raw Data (video and audio recordings) should be placed in a folder named rawData/, with three subfolders - labels/, RawAudio/ and RawVideo/.
The pickled audio and video convolutional neural networks named audioCNN13.pickle and videoCNN1.save should be placed in the cnn/ directory.
To run the tests, simply navigate to the root directory of ADOSAUR and run the following command:
py.test -v