setup.md

Setup

Install

• python2
• pip
• virtualenv with pip install virtualenv

Create a Virtual Environment

Run virtualenv -p path/to/python2 env. This will create a virtual environment where -p is the target interpreter for which to create a virtual (either absolute path or identifier string) (default: /usr/bin/python3) and will be saved to the env folder.

Activate the virtualenv with source env/bin/activate. The terminal should have (env) at the beginning. This indicates that the virtual enviroment is active.

Install dependencies with pip install -r requirements.txt

Add "image_dim_ordering": "th"and "backend": "theano" in your keras.json. This file should be in ~/.keras/ folder. This should look like:

{
    ...,
    "image_dim_ordering": "th",
    "backend": "theano"
}

Create a folder weights then download available pretrained models listed on README.md and save them in weights, the default is sam-resnet_salicon2017_weights.pkl. Make sure that line 109 is pointing to the correct .pkl file. Default should look like:

    m.load_weights('weights/sam-resnet_salicon2017_weights.pkl')

When done installing, the dev environment should be ready. Try running the command in README.md python main.py test path/to/images/folder/. The output should be in the predictions/ folder.

Exit the Virtual Environment

When done working on the project, deactivate the virtual environment by running deactivate on the terminal. To work on it again, just activate the virtual environment and skip all other steps.