EARhanced

Inspiration

Christine's grandmother is hard of hearing. Even with hearing aids, her grandmother often cannot hear, much less understand and participate in, conversations with her friends and family. This project was built in honor of her and to help people like her: elderly people who need a simple, intuitive interface to be able to communicate with those around them.

What it does

This project consists of a website, where a microphone is continuously on, converting speech to text. The results of the speech-to-text conversion is outputted in a text box with a large and easy-to-read font and a button to clear the text box if necessary. In addition, the website also supports multiple languages, including Spanish, French, English, and Chinese.

How we built it

This project was built only using React. Initially, we intended to build it using Flask and Python since speech-to-text seemed like a task more suited for Python, however, after finding out that React had a speech-to-text module that was free and outputted results more quickly, we switched to React.

Challenges we ran into

Because our web application continuously outputs results, we needed to have an algorithm that could convert speech-to-text in almost real-time. This challenge was difficult to solve, however, we researched various algorithms and Node package modules and tried our best to optimize the module we chose by specifying parameters and optimizing the code in React.

In addition, our user interface was initially build on Wix and Velo, but we found it difficult to integrate with our React code and functionality, so we had to rebuild the user interface all over again using HTML and CSS.

Accomplishments that we're proud of

We are proud of the fact that our website works and looks professional, is intuitive, and is easy-to-use.

What we learned

We learned about React and Node package modules that can expose functionality to turn speech-to-text. In addition, we also applied the CSS and HTML we learned in the workshop at VandyHacks to make our user interface.

What's next for EARhanced

In the future, we would like to implement a secure database that keeps track of conversations in this app, with the user's permission. We could then add a search bar to allow our user to search through their previous conversations. In addition, we would also like to add functionality to keep track of who said what using voice recognition software so that our output matches each person's words with their name. We look forwards to hosting this app online and helping the people who need the functionality it offers!

Built With

• css3
• html5
• javascript
• npm
• react
• react-native
• speech-to-text
• vercel