Wired Meetings is an online platform which aims to facilitate the online communication between blind and deaf specially during interviews by adopting a video chat app implemented on a website designed by us and also deployed on Azure with our domain. To use the app you have to signup and create a meeting then send the link of the meeting to the other person who also needs to signup.
The application developed using multiple frameworks:
- Backend implemented using
- NodeJs for web services
- Python, TensorFlow2 and OpenCV for Deep Learning models and Computer Vision Tasks
- Frontend implemented using basic CSS & React Framewrok
This Github Repo contains many directories:
- views : Contains the ejs files for chat, navbar and all the website pages
- public : Contains the CSS & designg files and images
- videos : Contains the recorded videos from the user during the meeting
- audios : Contains the recorded audios to and from the user during the meeting
- images : Contains the captured images from the user during the meeting
- models : Contains the ejs file which handles user authentication
- onnx-tensorflow : Contains ONNX library which is necessary for loading some of our Deep Learning models
- GP_Blind_Features : Contains the classes and trained models that supports our features for the Blind
- GP_Deaf_Features : Contains the classes and trained models that supports our features for the Deaf
You can find a demo video here
| Name | Contribution |
|---|---|
| Ahmed Mohammed Salah Abd El-Aziz | Blind Features Task Leader |
| Ibrahim Atef Abd El-Halim | Deaf Features Task Leader |
| Abd El-Hamid Amr Abd El-Hamid | Deployment on Azure Task Leader & Emotion Recognition |
| Mark Sameh Azer | Website Backend Task Leader & Azure Deployment |
| Fatma Elzahraa Mahmoud Esmail Mahmoud | Chatting system Task Leader & Website Frontend |
| Mariam Salah Abd El-Hamid | Website Frontend Task Leader & Chatting System |
More details about implementation and used model is in here
- Emotion Recognition
- Gender Classification
- Age Estimation
- Text to Speech
More details about implementation and used model is in here
- Sign Language Translator
- Speech to Text
Full integrated web platform supports:
- Login System that allows the user to register an account, login with his account and reset the password if it was ever forgotten
- Chatting room that allows two users to join the room and have a private conversation between each other
- Video and audio calling that allows the users within the chatting room to communicate with each other using audio and video not only messages
Deployed on Microsoft Azure with our domain with our reserved server so that anyone around the world can use our services
- No need to install anything just open our link and enjoy!
- Supported Browser : Chrome, Microsoft Edge, Mozilla Firefox...etc
We deployed our website on Microsoft Azure, Visit our link and enjoy our services:
http://wired-meetings.centralus.cloudapp.azure.com:3000/home
In the future we aim to support the following features on our platform ISA:
- Chatbot to help the user with his experience throughout the website and answer his/her questions
- Arabic Sign Language Translator instead of our American Sign Language Translator which we used due to the lack of the ARSL dataset