https://github.com/isseu/emotion-recognition-neural-networks

Emotion recognition using DNN with tensorflow
https://github.com/isseu/emotion-recognition-neural-networks

convolutional-neural-networks deep-neural-networks emotion-recognition machine-learning tensorflow

Last synced: 11 days ago
JSON representation

Emotion recognition using DNN with tensorflow

• Host: GitHub
• URL: https://github.com/isseu/emotion-recognition-neural-networks
• Owner: isseu
• License: mit
• Created: 2016-03-03T10:29:32.000Z (over 8 years ago)
• Default Branch: master
• Last Pushed: 2022-11-22T02:28:59.000Z (almost 2 years ago)
• Last Synced: 2024-08-01T17:23:30.752Z (3 months ago)
• Topics: convolutional-neural-networks, deep-neural-networks, emotion-recognition, machine-learning, tensorflow
• Language: Python
• Homepage:
• Size: 14.4 MB
• Stars: 836
• Watchers: 46
• Forks: 311
• Open Issues: 42
• Metadata Files:
  • Readme: README.md
  • License: LICENSE

Awesome Lists containing this project

• awesome-affective-computing - emotion-recognition-neural-networks
• awesome-affective-computing - emotion-recognition-neural-networks

README

        
# Emotion recognition with CNN

# DO NOT USE: Currently I think the code is not even running and I really don't have time to fix it

This repository is the out project about mood recognition using convolutional neural network for the course Seminar Neural Networks at TU Delft.

![Angry Test](https://raw.githubusercontent.com/isseu/emotion-recognition-neural-networks/master/paper/angry.png)

 67% Accuracy

 ![Angry Test](https://raw.githubusercontent.com/isseu/emotion-recognition-neural-networks/master/paper/matrix_final.png)

## Dataset

We use the [FER-2013 Faces Database](http://www.socsci.ru.nl:8180/RaFD2/RaFD?p=main), a set of 28,709 pictures of people displaying 7 emotional expressions (angry, disgusted, fearful, happy, sad, surprised and neutral). **The dataset quality and image diversity is not very good and you will probably get a model with bad accuracy in other applications!**

You have to request for access to the dataset or you can get it on [Kaggle](https://www.kaggle.com/c/challenges-in-representation-learning-facial-expression-recognition-challenge/data). Download `fer2013.tar.gz` and decompress `fer2013.csv` in the `./data` folder.

Install all the dependencies using `virtualenv`.

```bash

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.

```bash

$ python3 csv_to_numpy.py

```

By default this is using AlexNet architectures, but in the paper we propose different ones.

## Usage

```bash

# To train a model

$ python3 emotion_recognition.py train

# To use it live

$ python3 emotion_recognition.py poc

```

## Paper

[Link](https://github.com/isseu/emotion-recognition-neural-networks/blob/master/paper/Report_NN.pdf)