You can already slap together something with an RTSP stream and Frigate on an individual level. Some people have great views of the road, and some people could just count cars.
Frankly, I can see a lot of benefits for citizens if they could understand and verify which locations police enforce with their physical presence.
As an aside, where the public transparency and oversight for these schemes? It seems like such a powder keg.
> uses a Raspberry Pi 5, a Halo AI board, and You Only Look Once (YOLO) recognition software to build a “computer vision system that’s much more accurate than anything on the market for law enforcement” for $250
https://github.com/bennjordan/ALPRovingGround
> adversarial noise attacks on license plate recognition systems (see my PlateShapez demo).. create an output dataset to train more effective attack models.. small, hardly-noticeable, random gaussian shapes to confuse AI license plate readers
"Find Nearby Automated License Plate Readers", 70 comments, https://news.ycombinator.com/item?id=45487452
"Flock's gunshot detection microphones will start listening for human voices", 250 comments, https://news.ycombinator.com/item?id=45473698
"Tire Pressure Sensor IDs: Why, Where and When", 30 comments, https://news.ycombinator.com/item?id=45490202