← all entries
C-TIER repo

Open-source object tracking — the tech inside Flock cameras, free

billy.coder · 2026-08-08T23:44:36.000Z · in AI Saves, Vol. 8
Open post ↗

100k ft — What is this?

Open-source object tracking — the tech inside Flock cameras, free
repoC-tier

50k ft — Why care?

Repo Heist episode arguing the tracking behind $3k Flock cameras is free on GitHub: a 50k-star open-source library that detects, tracks and counts objects on any feed (doorbell, webcam, drone), demoed tracking every football player's speed, position and possession in one Python file. Repo name gated.
data scienceopen sourcesecurity

10k ft — Key takeaways

• Detection + tracking + counting is commodity open-source tech. • Sports analytics demos show the same stack.

5k ft — Tools & techniques

Python multi-object tracking, object detection

1k ft — Deep-dive / try it

Flock is going to be pissed someone Open sourced the blue print. Comment Repo and I’ll send it over
Transcript excerpt
A flat camera probably tracked your car today, your city paid 3 grand for it, and the tech inside? It's free on github. This is repo heist, episode 3, where I show you the best repos nobody's talking about. Flat cameras have been in headlines all year, lawsuits, cities canceling contracts, and the part nobody says out loud, the tech was never a secret. The tech, track, count, and it's all sitting on github for free. 50,000 stars, and all I do is point it at any feed, doorbell, webcam, drone, the football demo that you're watching below? Every player tracked at once, speed, position, possession, live, PSPN pays analysts for this, and you can do it all with one python file. Same tracking, your city rents for 3 grand, you can now set up for free. I'm sorry flocking had to do it, so if you want to test this out yourself, maybe on a home device, comment repo and I'll send it right over.

Collection: AI Saves, Vol. 8