eigenfaces

This example requires you to manually download the FERPlus and FER2013 csv dataset files.

More explicitly, you must extract the files "fer2013new.csv" and "fer2013.csv" from FERPlus and FER respectively, and put them inside the directory the directory ~/.keras/paz/datasets/FERPlus (if this directory is not there you must create it).

After that you must be able to run:

python eigenfaces.py

Alternatively you can extract the "fer2013.csv" and "fer2013new.csv" file and located them in a directory my_data and explicitly pass the args.data_path to eigenfaces.py i.e.

python eigenfaces.py --data_path my_data

To create database follow the following directory structure:

├── database
│   ├── images
│   │   ├── <person1_name>
│   │   ├── ├── image1.png
│   │   ├── ├── .
│   │   ├── ├── .
│   │   ├── <person2_name>
│   │   ├── ├── image1.png
│   │   ├── ├── .
│   │   ├── ├── .
│   │   ├── .
│   │   ├── .
│   ├── database.npy
├── experiments
│   ├── eigenfaces.npy
│   ├── eigenvalues.npy
│   ├── mean_face.npy
├── database.py
├── demo.py
├── eigenfaces.py
├── pipelines.py
├── processors.py