Our vision was to tie together sophisticated computer vision algorithms with the Meta SDK and headset in order to continuously keep track of the state of a blackjack game. In particular, we wanted the user to have access to a running count of the cards dealt in order to better know when to place bets. While some of the computer vision aspects proved hard to implement as well as we would have liked, we were satisfied with our full data pipeline setup. Images were constantly streamed from the camera to Python servers running in the background, which quickly computed the cards in the player’s vision to a pre-trained database and returned the most likely match.

Built With

Share this project:

Updates

Submission history