1 of 10

SF HACKS

Team Name: amul

Track : Health

​

Team members:

​

1. Pragya Moondra

2. Shreyansh Khandelwal

​

2 of 10

Problem Statement

In this pandemic struck world we all have shifted to Work From Home and Online classes culture, which for some have not been easy.

​

People, especially students are finding it difficult to cope with it and are falling prey to depression.

​

3 of 10

Our Proposed Solution:

​

​

We have created a deep learning, RNN model that examines the EEG report and classifies the emotion of a person as negative, positive or neutral.

If found negative, we can take further measures to help him/her.

4 of 10

Implementation

5 of 10

Dataset

  • We have used ‘EEG Brainwave Dataset: Feeling Emotions’ Dataset
  • The data was collected from two people using Muse EEG headband via dry electrodes
  • It has 2132 rows x 2549 columns

​

6 of 10

A sequence of Waveform from dataset

7 of 10

Accuracy

We have achieved over 95% of accuracy on the model, when trained on almost 1500 data-points.

8 of 10

Technologies Used

  1. Tensorflow-RNN model
  2. Python-Flask
  3. HTML
  4. CSS
  5. Notivize API

9 of 10

of project

Scope Of Project

​

  • Right now one needs to upload his/her EEG report converted into csv, to check the results.

​

  • This can be improved by converting the EEG report(waveform) into csv format on the server side and then giving results.

​

  • This can further be improved by integrating the model with the EEG hardware machine and give real-time predictions about the emotions of the person.

​

10 of 10

Thank you