NexTTS

Home Page 1
Home Page 2
Text-To-Speech Synthesis Area
Donate!
About Us
Our frontend and backend structures

Inspiration

Our interest in accessible text-to-speech began with our search for a model that would allow us to more effectively study written knowledge by listening to it. Currently, there are two types of easily accessible models available to the general public. The first type is the robotic-sounding AIs, which are essentially impossible to concentrate on. The others are very good but extremely expensive. OpenAI's text-to-speech model offers an affordable middle ground; however, it is not completely accessible to the general public due to the fact that it can only be used through its API. This inspired us to create NexTTS:

What it does

NexTTS is a front-end interface allowing easy use of the OpenAI model, which was previously inaccessible by non-coders. Its slick design allows you to effortlessly input text and promptly receive an audio file. It offers a variety of AI voices, allowing for all kinds of listening experiences. Our hope is that this affordable application will improve the studying experiences of students, especially those with visual disabilities, and therefore enhance their academic success.

How we built it

We built the front end using NextUI, a React-based library for website development. The backend was coded using Python to call the OpenAI API, and we used Flask to integrate those files together. Additionally, HTML and CSS were used to create a pleasant viewing experience.

Challenges we ran into

As we were very inexperienced with React, we often ran into issues with centering elements on the screen or having buttons placed where we wanted them to be. Other issues we had included linking up the backend Flask/Python with the frontend React. While the Wi-Fi was out, we also encountered many issues with debugging. It was not an easy project for us, but we learned a lot!

Accomplishments that we're proud of

We are proud that we were able to offer an affordable and higher quality API that is accessible to the general public. We had to learn much about React and Flask, and in the end, we had a fully functional website. We are happy that we were able to integrate everything together in one day's time. And we're proud of our very cool UI, made with Next-UI!

What we learned

Through the hackathon, we learned how to build a website from scratch using React and NextUI. Additionally, we improved our skills with various website-building languages (CSS, HTML, JavaScript) as well as Python. Additionally, we also learned how to integrate the OpenAI API with our program. Overall, we had a great experience at the hackathon and learned a lot, not only about coding but also about time management and dedication to the project. We also learnt that, apparently, the wifi at Marianopolis becomes a snail after midnight.

What's next for NexTTS

This is just the beginning for NexTTS! We plan to greatly expand the capabilities of the website, introducing compatibility with PDFs and allowing for other text-to-speech models to be used. We would love to have our application used by more students who, like us, are also tired of reading from a screen and are in search of a cheap text-to-speech extension. Additionally, if enough users find our website and use it, we will have to add advertisements or ask for payment due to the cost of usage. Overall, we are incredibly excited for the future of this project and its potential to help students around the world.