Emotion Enhanced Chat

EmoChat is the 1st place winner of a 2-days hackthon named Brainihack. This project is based on OpenTokRTC and enhances with emotion detection capabilities using NeuroSteer EEG headset.

The idea is to detect the user emotion using live EEG feed and enable emotion sharing over the chat room by gestures like emojis next to chat line and color changes according to peer's mode.

File Overview

• Procfile is required to run the nodejs app on Heroku
• package.json contains all npm modules to run the app
• app.js contains all server side code
• config.js contains configurations: TokBox credentials, p2p mesh support, Redis support, reserved rooms, etc.
• lib folder contains all the code to handle configurations: p2p mesh support, Redis support, reserved rooms, etc.
• views folder contains the html template for the app
• public/css folder contains all the css for the app.
  Look for files with .scss extensions. .css files are generated from sass.
• public/js contains the front end code and interactions with OpenTok SDK.

How to run the app:

1. Clone this repo
2. Get my API Key and Secret from TokBox
3. Replace OTKEY and OTSECRET with your corresponding API Key and Secret in config.js or by setting your env variables
4. Run npm install to install the necessary packages
5. Start the server with node app.js

How to use:

1. In the chat main screen (usaly localhost:9393) enter a room name (anything would do)
2. type in the chat /help to get a list of configurations available
3. To link the NeuroSteer headset with the current chat user type /user <last 6 digits of the headset serial #>

Additional links:

1. https://vimeo.com/122224388 - presentation at Brainihack final demo
2. https://vimeo.com/122826001 - screen capture of a chat we did in Brainhack's final demo