jonathanloganmoran / emotion-recognition-neural-networks

A convolutional neural network for human emotion classification

Emotion recognition with CNN

Contributing author: @jonathanloganmoran

Computational Cognitive Neuroscience (CSE 173) | Spring '18 | Prof. David Noelle

Forked model used in a Seminar Neural Networks course at TU Delft, published by @isseu

67% Accuracy

Classification benchmarks via @isseu

Dataset

We use the FER-2013 Faces Database, a set of 28,709 pictures of people displaying 7 emotional expressions (angry, disgusted, fearful, happy, sad, surprised and neutral).

You have to request for access to the dataset or you can get it on Kaggle. Download fer2013.tar.gz and decompress fer2013.csv in the ./data folder.

Install all the dependencies using virtualenv.

virtualenv -p python3 ./
source ./bin/activate
pip install -r requirements.txt

The data is in CSV and we need to transform it using the script csv_to_numpy.py that generates the image and label data in the data folder.

$ python3 csv_to_numpy.py

Usage

# To train a model
$ python3 emotion_recognition.py train
# To use it live
$ python3 emotion_recognition.py poc

Presentation/project overview

Link

Original Paper via @isseu

Link