All the products being developed today with AI technologies should be utilized by each one of us equally. The biggest motivation for this project was to give back something to society by using the provided resources and trending data science techniques most efficiently.

What it does

This project converts a sign language video to a text which is echoed by the speaker of the computer. This query is then sent to Alexa/Google Mini for a response. Then the response is converted back to a sign language video.

How we built it

We used the following techniques and products:

1) Google Cloud Speech Recognition API - To convert Alexa/Google Mini speech to text and vice-versa. 2) Deep Learning Models(Convolution Neural Networks) - To train our model to interpret the video of sign language and convert it to text. 3) HTML/Javascript - Front End Development 4) Python/Tensorflow - Backend for backend development 5) Flask - Connecting Backend and Front End 6) Alexa/Google Mini - Quering 7) Github - Repository for our Code

Challenges we ran into

1) Dataset for Sign Language 2) Training of huge images 3) The accuracy of the Model within a short span of time

Accomplishments that we're proud of

Building an end-to-end product for society.

What we learned

1) Working as a team 2) Handling Pressure 3) Have fun while working

What's next for Helping Hands

1) Live Video Streaming as an Input 2) Training on a larger dataset to improve accuracy

Built With

Share this project: