Emotion-Detection---MAIS-Project

IMPORTANT Before running, add the New_Model.h5 file to the folder. The file can be found here. https://drive.google.com/file/d/1h2p8sYgSynF1H00DCOaQTABwuh-P9tOT/view?usp=sharing

About the Project

Emobot uses a Machine learning model for facial and emotion recognition. The model was trained using the AlexNet CNN with preprocessed grayscale images using Tensorflow and Keras. It was then deployed onto a webapp using Flask. The app is able to detect faces and predict emotions in realtime and display the emoji and text corresponding to the emotion.

Data

The following dataset was used to train our model. It contains grayscale images of 7 different emotions. https://www.kaggle.com/datasets/ananthu017/emotion-detection-fer

Organization

1. Code Contains ipynb files for our training results
2. static CSS file for landing page
3. templates HTML template for landing page
4. video.py python code for the videofeed and to predict using the model.
5. app.py main python script to instantiate Flask server
6. Pdfs and text files Project Proposal.txt, Preliminary Results and Final Results are also included for updates on the project.