Inspiration

Providing accessible means of media consumption is not only a legal requirement, but a moral obligation. And not only in the quickest way possible, but also in the easiest way possible, even if it is more challenging for the developers. With current advancements in deep-learning technologies, applications and hardware, providing easy translation and transcription services, especially in scenarios where having a translator may not be ideal (live events, especially with large crowds, will certainly improve the quality of life of the deaf community.

What it does

The app works in many capacities, but at its core, it uses the devices microphone to identify and translate the speech received, using Deepgram's API, and display it in an easily readable fashion on the users computer or phone/tablet.

How we built it

Deepgram API, Flask backend, simple frontend technologies, hosted on Google Cloud's APp Engine

Challenges we ran into

Developing Flask backend securely, integrating with Google Cloud, mobile app development

Accomplishments that we're proud of

Solving challenges above

What we learned

Java for mobile apps (android), Flask backend, google cloud.

What's next for Accessible Speech to Text Generator and Translator

Developing 3D model to translate to sign language as well, if preferred (through user surveys). Integrate into wearable devices (i.e. smartwatch, and possibly smart glasses).

Share this project:

Updates