Digitally tag different video takes with slate info (eg. description, scene) and identifying actors for easy preview.
We are filmmakers who understand how difficult it is to preview videos of different takes (sometimes up to hundreds of videos).
What it does
Trans Slate recognizes text on a handwritten/typed slate and digitally transcribes it into a graphic overlay on the thumbnails of each video. It also detects faces for further identification of talents and recognizes voice input of the quality of each take (eg. excellent, good, no good).
How I built it
We used Ziggeo API to create a recording app. We then detect frames that contain the slate information and transcribed it to text using Google Cloud API. Lastly, we built an image set of actors in the video and used Clarifai’s API to detect actors.
Challenges I ran into
Getting the different API to integrate together to work.
Accomplishments that I'm proud of
Getting each API to work each step along the way.
What I learned
How to implement various APIs
What's next for Trans Slate
Reaching out to non-profits/government institutions and small businesses/solopreneurs who tend to have limited resources to engage professional filmmakers to produce marketing materials. Trans Slate is a simple and elegant solution to saving time and money during the production process.