Inspiration

The pandemic hits hard on the world, We all are worried about our health. It was important to check heart rate, mental health, and detect any persistent coughing. We need to get live status of active cases near our locality so that we can plan our journey accordingly. For this purpose I just connected with Dr. Anjali and started working on this project.

What it does

You can ask question without clicking a button or typing. Just ask and she will give your health-status, how to care your children, active cases near you and much more. For active cases we used the covid19 India API.

How we built it

There are 4 version for this assistant

  1. Build on Nod-red
  2. Build on Flask web app
  3. Build on Android
  4. Deployed on Nvidia Jetson Nano/PC with automated cough detection, response to your voice using RNN based Noise-reduction technology

Challenges we ran into

It has taken tremendous effort to build this during the hackathon. Including training multiple AI model in limited amount of time, Configuring IBM cloud, Node-red application, Building a flask RESTful app and android application. Installing dependencies on Nvidia Jetson Nano and configuring the script on this Single Board Computer(SBC) has taken so much of our time.

Accomplishments that we're proud of

Succesfully build the system working on SBC's such as Nvidia Jetson Nano, which acts as a local brain for the Asssistant.

What we learned

Training Deep learning models, IBM-cloud devops, AI edge device integration with cloud.

What's next for Deep learning Covid-19 chatbot

Now we are almost completed our work for integrating the emotion recognition, heart rate detection over the cloud. so that it can be used as an API call. which helps in integrating on mobile applications, web application,IoT devices, Other SBC's

Built With

Share this project:

Updates