brianlevis / nerf-iron-dome Goto Github PK

Maybe update the readme with a few links?

A few folks have simplified opencv, and such, updated learning models, etc. If it's dead for now (your work) a few refs might help others.

Jeff feeling does opencv on rpi4, and hits the 2fps barrier, but makes leaf lower + facial detection
https://www.jeffgeerling.com/blog/2021/face-detection-my-leaf-blower

Or go nuts with an external tpu unit on the USB, and tensorflow it up from there
https://bestofcpp.com/repo/NobuoTsukamoto-edge_tpu-cpp-artificial-intelligence

(Which honestly if you can crank out good c++, and have $60 to spare for the unit, seems best. Again, sample size data set/pickle is essential)

Tpu edge unit
https://coral.ai/products/accelerator/

Hope this helps someone who stumbles here! :)

Seemed feature complete... but would love to get some schematics and parts list for everything. I have everything else ready to go.

You did a good think making this.

Is this still being worked on, before I try to work from is again,

One day, I might just go with the idea of a tensorflow based system, which the RPI4 8gb will still likely choke on horribly, just as it just with OPencv/CV2?

( (Note for other who read this pr: Don't ask when) it's fall in Boston, maybe I'll write it when it's super late in the winter..) )

(If I do a new thing, it would be announced on reddits nerfmods group, or on fbook https://www.facebook.com/groups/164130663628549 )