This project aims at determing facial emotions from cell phones at the retail store. The project deals with using multiple available datasets such as KDEF, RAF and FER2013 for emotion recognition. There are slight changes needed with RAF to load the data. The project uses Convolutional Neural Network with Dense Layer to determine the emotion of the person. It uses VGG 19 as the base model.
git clone https://github.com/Tabish06/FERRA.git
Copy shape predictor file to cloned folder: https://drive.google.com/file/d/1blALLnjbpvyHHi8xOZmHPUgna0E5Jurb/view?usp=sharing
Note that this will require Python. To install Tensorflow using pip, run the following:
pip install --upgrade tensorflow
To verify the install:
python -c "import tensorflow as tf;"
Installing using PyPI:
pip install keras
To train the mode, run python evalute.py