Mirrors/blakeblackshear.frigate
1
0
Fork 0
mirror of https://github.com/blakeblackshear/frigate.git synced 2026-02-20 13:54:36 +01:00
Code Issues Projects Releases Wiki Activity

Miscellaneous Fixes (#20897)

Browse Source 
* don't flatten the search result cache when updating

this would cause an infinite swr fetch if something was mutated and then fetch was called again

* Properly sort keys for recording summary in StorageMetrics

* tracked object description box tweaks

* Remove ability to right click on elements inside of face popup

* Update reprocess message

* don't show object track until video metadata is loaded

* fix blue line height calc for in progress events

* Use timeline tab by default for notifications but add a query arg for customization

* Try and improve notification opening behavior

* Reduce review item buffering behavior

* ensure logging config is passed to camera capture and tracker processes

* ensure on demand recording stops when browser closes

* improve active line progress height with resize observer

* remove icons and duplicate find similar link in explore context menu

* fix for initial broken image when creating trigger from explore

* display friendly names for triggers in toasts

* lpr and triggers docs updates

* remove icons from dropdowns in face and classification

* fix comma dangle linter issue

* re-add incorrectly removed face library button icons

* fix sidebar nav links on < 768px desktop layout

* allow text to wrap on mark as reviewed button

* match exact pixels

* clarify LPR docs

---------

Co-authored-by: Nicolas Mowen <nickmowen213@gmail.com>
This commit is contained in:
Josh Hawkins
2025-11-17 08:12:05 -06:00
committed by GitHub
parent 097673b845
commit fbf4388b37
23 changed files with 321 additions and 220 deletions
Download Patch File Download Diff File Expand all files Collapse all files

8
frigate/data_processing/post/review_descriptions.py
View File

@@ -132,17 +132,15 @@ class ReviewDescriptionProcessor(PostProcessorApi):
if image_source == ImageSourceEnum.recordings:
duration = final_data["end_time"] - final_data["start_time"]
buffer_extension = min(
10, max(2, duration * RECORDING_BUFFER_EXTENSION_PERCENT)
)
buffer_extension = min(5, duration * RECORDING_BUFFER_EXTENSION_PERCENT)
# Ensure minimum total duration for short review items
# This provides better context for brief events
total_duration = duration + (2 * buffer_extension)
if total_duration < MIN_RECORDING_DURATION:
# Expand buffer to reach minimum duration, still respecting max of 10s per side
# Expand buffer to reach minimum duration, still respecting max of 5s per side
additional_buffer_per_side = (MIN_RECORDING_DURATION - duration) / 2
buffer_extension = min(10, additional_buffer_per_side)
buffer_extension = min(5, additional_buffer_per_side)
thumbs = self.get_recording_frames(
camera,

2
frigate/data_processing/real_time/face.py
View File

@@ -424,7 +424,7 @@ class FaceRealTimeProcessor(RealTimeProcessorApi):
if not res:
return {
"message": "No face was recognized.",
"message": "Model is still training, please try again in a few moments.",
"success": False,
}