Video is a great form of media, but it's often to slow and too long. We want a way to find the most important parts of a video, and only watch those parts.

What it does

Users can submit a video Scribr, and Scribr will automatically find the most important parts of the video, and will provide a brand new video, trimmed to just the important moments.

How we built it

Scribr first transcribes the video using the RevSpeech API. Then it does a form of sentiment analysis on the transcribed videos, using NLP, to find the most important sentences. It then finds the timestamps of all of the important videos, and then re-cuts the video to trim down to only those sentences.

Challenges we ran into

So many challenges. From installing Python on a windows computer (for a brand new coder!) to automatically trimming videos on a server... we were genuinely shocked the first time the entire process worked, and we played a trimmed video that actually worked.

It was so seamless, we actually wasn't sure if it worked the first time.

Accomplishments that we're proud of

We are proud that we came together with a diverse skillset, including someone who just learned how to code, and were able to create a real, working system.

What we learned

One of us learned to code. The others learned Flask, RevSpeech, how a computationally intense backend system worked, how to work quickly, and how to power through until the end.

What's next for Team Two

I say dance, they say 'How high?' We're got many more ideas that we want to add to this project. And as soon as we sleep for a week or two, we'll get back to adding them.

Share this project: