Inspiration
We often have trouble coming up with ideas when preparing presentations. Also, students and teachers cannot give presentations and lectures when they are feeling ill or are absent, without going through a lengthy recording process—even worse if they have a sore throat.
What it does
EZ-Present takes your slide decks and turns them into fully-voiced video presentations with your own voice. These can be used to make up for your absence, improve your presentation skills, provide inspiration when preparing a presentation, create digital lessons, and much more.
How we built it
EZ-Present is built with Next.js handling the frontend and backend, MongoDB for persistence.
A slide deck is provided to Gemini, which generates a transcript from the slides. The transcript is passed to Fish Audio, which generates TTS.
The TTS audio is passed again to Fish Audio to generate a transcript with timestamps, these are used to synchronize the audio with video.
Finally, a slideshow is generated with FFmpeg and combined with the TTS audio.
Challenges we ran into
Synchronizing slide transitions in the transcript with the video presentation.
Accomplishments that we're proud of
Generating TTS audio Fish Audio.
Exploring a unique idea using AI.
What we learned
How to use AI for automation.
Streaming files on the web.
What's next for EZ-Present
Improved frontend.
Full slideshow generation.
Built With
- fishaudio
- mongodb
- next.js
