@NickM-27 commented on GitHub (Oct 4, 2023):
this is by design, if you only want motion recorded then that is all that will be kept. if motion is not occurring then there is nothing to save for…
@NickM-27 commented on GitHub (Oct 5, 2023):
This is working as expected. Zone level filters are used for affecting if that object has entered the zone or not. Only the camera level filters…
@NickM-27 commented on GitHub (Oct 4, 2023):
this may be some network issue, I've never had one fail and haven't seen other reports like this either. Blake should be able to check on the frigate+…
@ryosoftware commented on GitHub (Nov 1, 2023):
For now I'm stay on not custom model due to the custom model has too many false positives. Also I've changed my server host to another model. Due…
@NickM-27 commented on GitHub (Oct 31, 2023):
Is this still happening or can we close this?
@NickM-27 commented on GitHub (Oct 4, 2023):
what I am saying is I don't see any issues related to the frigate+ model, all of the processes that the object detection model would be involved in…
@ryosoftware commented on GitHub (Oct 4, 2023):
I apologize, but I do not understand to what model do you refer or what screenshot do you need
@NickM-27 commented on GitHub (Oct 4, 2023):
sorry what I mean is, which model were you using at the time of the screenshot (the default or frigate+). I am not asking for, nor do I have access to…
@CurseStaff commented on GitHub (Oct 4, 2023):
Your movement_weights numbers look a little strange to me as well.
Who makes you say that, I'm curious to know :)
@Majestic7979 commented on GitHub (Oct 12, 2023):
Indeed, it was my port. Facepalm... embarrassing! I changed the external port of my Frigate docker container. Now it is working... apologies for…
@ryosoftware commented on GitHub (Oct 4, 2023):
Frigate+
@hawkeye217 commented on GitHub (Oct 4, 2023):
In testing with my particular camera (a lower end Dahua) I've found that the predicted time values from a simple linear regression are not accurate…
@NickM-27 commented on GitHub (Oct 4, 2023):
Based on my understanding, there is too much ambiguity in the code differences to say whether or not the code changes make a difference, especially…
@NickM-27 commented on GitHub (Oct 4, 2023):
yeah the CPU usage is all related to ffmpeg / stream capture, nothing to do with object or motion detection