AudioAlpha

Inspiration

We were inspired to make this project by knowing from experience how difficult it is to make a presentation, especially when you do not have a parent or anyone else watching to gives critiques. With this application, one can prepare for a presentation without the help of a parent or anyone else watching. This makes it easy to find exactly where one messes up on a presentation as well as ways to fix these issues.

What it does

This project uses various features to rate and critique a person's presentation. One feature is that it finds pauses in audio, or what the presenter is saying. This allows them to know when they pause, or where they make mistakes or are unsure of what to say. Another feature is that it finds repeated words, which also allows them to find mistakes.

How I built it

We used Python, HTML, and Javascript to make this project. The project is built on a web application, but it uses Python in order to analyze the audio file and give critiques.

Challenges I ran into

One challenge that we ran into was the lack of free API services for speech-to-text. Speech-to-text is such a useful feature, but it is almost impossible to create a decently accurate speech-to-text service in one day, making it essential to find a good free one. We spent time trying to implement several such as Google's and Microsoft's, but only had success once we found the Google Web Speech-to-Text.

Accomplishments that I'm proud of

We are proud of creating a successful project within 24 hours using Python that serves a great purpose to people that may struggle in presentations. In addition, we challenged ourselves by choosing a project that involves audio when having little experience with analyzing audio.

What I learned

We learned to work as a team in collaborating in such a large project as well as analyzing voice. We did not have knowledge on voice analysis, but we decided to take on the challenge and learn through experience.

What's next for AudioAlpha

We plan to polish AudioAlpha more and add more features. Although we only had 24 hours, we got some of the core features done, and we feel that it serves well as a project. The interface could use some work, but we feel that the most important part as the core features.