@jornodberg commented on GitHub (Nov 23, 2023):
Wow, what a fast respons! Thanks a lot.
Could you elaborate how I could accomplish this using websocket? I have no experience with websocket,…
@NickM-27 commented on GitHub (Nov 23, 2023):
MQTT would be the easiest to do https://docs.frigate.video/integrations/mqtt along with some MQTT tool that has a CLI
@AlasdairH commented on GitHub (Nov 23, 2023):
Thanks for clearing that up, makes sense. I'm getting myself set up to make some contributions to the repo, starting small with some tooltip clarity…
@blakeblackshear commented on GitHub (Nov 23, 2023):
They are both correct, just different things. The score in the image is the score from that single frame. The score on the event is the top…
@NickM-27 commented on GitHub (Nov 23, 2023):
The model can already be defined at the detector level. However, multiple detector types are not supported as that is not as simple as it seems.
@NickM-27 commented on GitHub (Nov 23, 2023):
This is possible via MQTT and via websocket.
@StuartRoberts40 commented on GitHub (Nov 15, 2024):
Sorry to resurrect this one but I am struggling. I have read many different guides on setting up Frigate with a Raspberry Pi and my IP camera…
@NickM-27 commented on GitHub (Nov 23, 2023):
I just realized you're running detect at 1280x720 and that's likely fine though you could reduce CPU usage some if detect was run on the reolink ext…
@fisch55 commented on GitHub (Nov 24, 2023):
sorry one more question... do I need this Lines? What do you think?
environment_vars: LIBVA_DRIVER_NAME: iHD
@NickM-27 commented on GitHub (Nov 22, 2023):
I'd suggest reading https://docs.frigate.video/frigate/camera_setup#choosing-a-detect-resolution