D'SpeakerVerse

Inspiration

Our inspiration for D'SpeakerVerse stemmed mainly from our personal experiences. We have in many ways encountered challenges with using speech-to-text AI like Google Translate and Alexa because of our accents. Also, we discovered that people who do not have English as their first language find it hard to follow through movies and TV shows without subtitles. We decided to integrate some cultural aspects in future developments of D'SpeakerVerse as this will make people feel home away from home.

What it does

An app, it's taking input from the user so it should be able to correct speech from text. If you take speech from the user and process it in its memory, so since we're doing tongue twister game it should relay back the speech in the same way. It should try to mimic the speaker's voice, a streak: when the user signs in for 3 days in a row they keep adding onto their streak. When the user first installs the app, they are asked to sign up, and then they'll introduce themselves to the app (the app will ask questions about their age, ethnicity, and gender). And from that date, the app will be able to customize: the appearance and maybe the audio.

How we built it

We used Figma to build our UI and our design, outlining our proof of concept. In addition, we used Flutter and Dart to build our application, making it cross-platform compatible with Android and iOS, which will be able to help us reach more end users. Finally, we used Firebase for our database and Google OAuth for the authentication system.

Challenges we ran into

We ran into the challenge of creating a solution to a problem that was unique and not already done. We also ran into trouble understanding a way to ensure scalability and inclusivity of all minority groups. We started with brainstorming various ideas and built off of those ideas to come up with specific solutions to problems.

Accomplishments that we're proud of

Some accomplishments that we're proud of us is that we made a functioning application within 24 hours with our main features pin-pointed and that we learned a new software very quickly. Also, all of us had little to no experience with machine learning, and after attending workshops and taking a quick course on it we have improved our learning on what it is. Along with the fact we came up with a pretty amazing idea that could benefit others including ourselves.

What we learned

We learned a lot including what Figma is and how to use it, we also learned about many different kinds of applications used for programming and how AI affects our daily lives. We also learned a lot more about so many problems that we could've never imagined and how there are a lot of solutions out there that aren't implemented.

What's next for D'SpeakerVerse

Our next steps are to incorporate different features with the intent of improving the experience of users and allowing for it to become more broadly applicable to many different scenarios. Some of our future goals include Audio Readings, Criminal Applications (after extensive research has been done and a data set has been collected), Research Applications, Character Separation( to include different voices for different characters for audiobook readings), and Coaching Services ( to aid the user with mastering pronunciations in their accent.