Inspiration

We were inspired to create a program that would convert ASL (American Sign Language) to text or voice commands, in order to simplify communication. In a world full of constantly-evolving technology, we are always being shown examples of smart devices that are voice-activated, but devices like these may not be accessible and inclusive for everyone & we don't want this disability to stop people from achieving their goals! That's why we created Speak Out!.

What it does

Speak Out! is a program that uses computer vision and artificial intelligence to convert ASL signs into English text or audio, and vice versa. It takes input from a user who finger-spells letters, and the phrase is then interpreted into text or audio.

How we built it

We built Speak Out! using Google Teachable Machine and the interactive dashboard using HTML,CSS,Javascript & Netlify. As we used Teachable it allowed us to concentrate a lot more on the data to train to give out more accurate results!

Challenges we ran into

Initially we decided to finish the project with TensorFlow and integrate with Google Cloud! But we ran into a lot of problems with TensorFlow so there was no time to debug & it was too late to shift to OpenCV to train the models. Hence we shifted to no-code platforms - Google Teachable Machine! and our results speaks now! ;)

Accomplishments that we're proud of

As stated earlier we ran into unexpected and major problems and we are really proud that we didn't give up and made some very important We are proud that we didn't panic but instead took quick on-the-spot decisions! We were able to produce a finished product despite all of the technical difficulties we experienced. However, we also learned a lot. We learned more about using computer vision and AI, training models etc. For some of us, this was also our first hackathon, and we are proud of what we created in such a short time.

What's next for BitBots

If we had more time to work on this project, we would have loved to continue improving the accessibility of the app, by adding in functionalities to convert text or audio to ASL. We know that there is still a lot of progress to be made toward making tech accessible for everyone, and we hope to continue to work on this in the future. We also look to convert the current functionality to a revenue based model for corporate for business and shareholders meeting and the main functionality in the education sector for students with disablities! We look forward to make our project into a complete LaaS (language as a Service)! Cheers!

Built With

Share this project:

Updates