@NickM-27 commented on GitHub (Feb 5, 2026):
This would be very involved, basically changing the internal presets via MQTT which would require restarting all ffmpeg streams related to that…
@ZhaiSoul commented on GitHub (Feb 5, 2026):
If you need to know the progress of file deletion, you first have to traverse all the files, get their sizes, and then calculate the time required. In…
@hawkeye217 commented on GitHub (Feb 3, 2026):
This is already supported when using a model (either a custom model or Frigate+) that natively detects license_plate objects. As indicated by the…
@blakeblackshear commented on GitHub (Feb 4, 2026):
Requested
@Ivan2getdmodelz commented on GitHub (Feb 3, 2026):
I read your documentation, and read that it requires training a general model to output a license plate as a detection., or paying for a…
@hawkeye217 commented on GitHub (Feb 3, 2026):
We could potentially expose the plate's coordinates via the tracked_object_update MQTT topic, but to draw a bounding box using those coordinates…
@santibur06 commented on GitHub (Feb 3, 2026):
Hi Nick,
While what you mention is true, I’m specifically referring to the main camera grid view where all cameras are displayed.
Camera groups…
@NickM-27 commented on GitHub (Feb 2, 2026):
Batching requires model support, all models used in Frigate currently don't support a dynamic input size.
@NickM-27 commented on GitHub (Feb 3, 2026):
That would likely just be made easier when the full config is added to the UI, camera groups is how cameras are customized including streaming…