Prize goal

Best freshman hack Most socially useful hack


Transcribing subtitles is a pain in the ass

What it does

In goes your raw video, out comes subtitled video

How I built it

Mainly I leverage on Google Speech API (Speech Recognition) to detect words from the audio source. I use fluent-ffmpeg to manipulate the video data (getting the audio from video, merging the subtitles with the video).

Challenges I ran into

I am a freshmen with no experience in using the Speech API nor the ffmpeg library. I came with an idea without any knowledge on how to execute it. So I had to explore the options and just try.

Also callback hell is still a thing because ffmpeg doesnt use promises :(

...Also, recording a meta presentation is a bit tough.

Accomplishments that I'm proud of

This is my first ever solo hackathon and I managed to complete the project on my own. So that's substantial to me. I joined this hackathon with the intention to win. So if I do, that would be awesome!

What I learned

I learned that I am actually happier and more productive going solo because this idea is freshly my own and not from a group. So seeing it comes to fruition is best.

Besides it, I learnt a bit of machine learning from the workshop. I am quite clear now how Google Speech API works as well as ffmpeg.

What's next for Video Auto Captioner

I had the idea to make a webpage for it but that means users will have to upload the video to my server, so I think it is better to keep it as a script file where only the audio is uploaded to the Speech API.

Built With

Share this project: