johanels/photoprism
1
0
Fork 0
mirror of https://github.com/photoprism/photoprism.git synced 2025-12-12 00:34:13 +01:00
Code Issues Packages Projects Releases Wiki Activity

AI: Update face flag notes in internal/config/report_sections.go #5167

Browse Source 
Signed-off-by: Michael Mayer <michael@photoprism.app>
This commit is contained in:
Michael Mayer
2025-11-16 10:57:45 +01:00
parent 6438b70d70
commit 4a0f9bfa7a
1 changed files with 1 additions and 1 deletions
Download Patch File Download Diff File Expand all files Collapse all files

2
internal/config/report_sections.go
View File

@@ -9,7 +9,7 @@ type ReportSection struct {
// see https://docs.photoprism.app/getting-started/config-options/#face-recognition
var faceFlagsInfo = `!!! info ""
To [recognize faces](https://docs.photoprism.app/user-guide/organize/people/), PhotoPrism first extracts crops from your images using a [library](https://github.com/esimov/pigo) based on [pixel intensity comparisons](https://dl.photoprism.app/pdf/publications/20140820-Pixel_Intensity_Comparisons.pdf). These are then fed into TensorFlow to compute [512-dimensional vectors](https://dl.photoprism.app/pdf/publications/20150101-FaceNet.pdf) for characterization. In the final step, the [DBSCAN algorithm](https://en.wikipedia.org/wiki/DBSCAN) attempts to cluster these so-called face embeddings, so they can be matched to persons with just a few clicks. A reasonable range for the similarity distance between face embeddings is between 0.60 and 0.70, with a higher value being more aggressive and leading to larger clusters with more false positives. To cluster a smaller number of faces, you can reduce the core to 3 or 2 similar faces. It is **strongly recommended** that you run the "photoprism faces reset" command in a terminal to remove existing clusters and mappings after changing any of the clustering parameters, as otherwise inconsistencies may result in unexpected behavior or errors.
A reasonable range for the similarity distance is between 0.60 and 0.85, with higher values resulting in more aggressive clustering and more false positives. To cluster a smaller number of faces, reduce the core to 3 or 2 similar faces. After changing any of the clustering parameters, it is **strongly recommended** that you run the "photoprism faces reset" command in a terminal to remove existing clusters and mappings, as otherwise inconsistencies may result in unexpected behavior or errors.
We recommend that only advanced users change these parameters:`