administrator/frigate
1
0
Fork 0
mirror of https://github.com/blakeblackshear/frigate.git synced 2026-05-04 04:27:42 +03:00
Code Issues Actions 2 Packages Projects Releases Wiki Activity

Clarify face recognition docs

Browse Source
This commit is contained in:
Nicolas Mowen 2025-05-14 14:33:46 -06:00
parent c2e28fbc69
commit 65b6d50c76
1 changed files with 7 additions and 5 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

12
docs/docs/configuration/face_recognition.md
View File

@ -105,19 +105,21 @@ When choosing images to include in the face training set it is recommended to al
::: :::
### Understanding the Train Tab
The Train tab on the face library is used to see recent face recognition attempts. The face images are grouped by the person object that they were detected as being related to. Under each face image there will be a name (or `Unknown`) and the confidence of that recognition. Each image can be used to train as a specific person, but not every image should be trained. See below for guidance on training.
### Step 1 - Building a Strong Foundation ### Step 1 - Building a Strong Foundation
When first enabling face recognition it is important to build a foundation of strong images. It is recommended to start by uploading 1-5 photos containing just this person's face. It is important that the person's face in the photo is front-facing and not turned, this will ensure a good starting point. When first enabling face recognition it is important to build a foundation of strong images. It is recommended to start by uploading 1-5 photos containing just this person's face. It is important that the person's face in the photo is front-facing and not turned, this will ensure a good starting point.
Then it is recommended to use the `Face Library` tab in Frigate to select and train images for each person as they are detected. When building a strong foundation it is strongly recommended to only train on images that are front-facing. Ignore images from cameras that recognize faces from an angle. Then it is recommended to use the `Face Library` tab in Frigate to select and train images for each person as they are detected. When building a strong foundation it is strongly recommended to only train on images that are front-facing. Ignore images from cameras that recognize faces from an angle. Aim to strike a balance between the quality of images while also having a range of conditions (day / night, different weather conditions, different times of day, etc.) in order to have diversity in the images used for each person and not have over-fitting.
Aim to strike a balance between the quality of images while also having a range of conditions (day / night, different weather conditions, different times of day, etc.) in order to have diversity in the images used for each person and not have over-fitting. You do not want to train images that are 90%+ as these are already being confidently recognized. In this step the goal is to train on clear, lower scoring front-facing images until the majority of front-facing images for a given person are consistently recognized correctly. Then it is time to move on to step 2.
Once a person starts to be consistently recognized correctly on images that are front-facing, it is time to move on to the next step.
### Step 2 - Expanding The Dataset ### Step 2 - Expanding The Dataset
Once front-facing images are performing well, start choosing slightly off-angle images to include for training. It is important to still choose images where enough face detail is visible to recognize someone. Once front-facing images are performing well, start choosing slightly off-angle images to include for training. It is important to still choose images where enough face detail is visible to recognize someone, and you still only want to train on images that score lower.
## FAQ ## FAQ