Skip to content

Student project of e-Health system. Real-time recognition of emotions and pulse detection from webcam video.

License

Notifications You must be signed in to change notification settings

Carlloy/Recognition-of-emotions-using-CNN

Repository files navigation

Recognition-of-emotions-using-CNN

Student project of e-Health system. Real-time recognition of emotions and pulse detection from webcam video. More information about this project is in file scientific_paper_about_project.pdf. Till now it is only written in polish language.

Dataset

We are not allowed to publish FER2013 dataset. If you want to train model then you should download it on your own from Kaggle. Download fer2013.tar.gz from this website and decompress fer2013.csv in the ./training/dataset folder.

Requirements

To install all dependecies you will need pip.

pip3 install -r requirements.txt

Usage

These are the commands that will allow you to start the program. The first one is used to train the model. You will need to download the FER2013 dataset before training. The second command is used to run an application used to detect pulse and emotions from the webcam image. Run it from the directory of this project.

# To train model
$ python3 -m training.train_model.py
# To use the app
$ python3 -m  application.app.py

About

Student project of e-Health system. Real-time recognition of emotions and pulse detection from webcam video.

Resources

License

Stars

Watchers

Forks

Releases

No releases published

Packages

 
 
 

Languages