face clustering minimum

[31337-4554551n]

Face recognition: 0 Queued files, Last classification: 12 hours ago, Scheduled background jobs: 0, 

 Face clustering: 42 faces left to cluster, Last clustering run: 12 hours ago, Scheduled background jobs: 0,
A minimum of 120 faces per user is necessary for clustering to kick in

If I'm reading this correctly, I have 78 pics with faces detected, but they won't be added into their correct people until I take another 42 photos with faces in them?

Is there any way to shrink those batches?
I definitely know I have photos with peoples faces that aren't showing up under people... and idk how to force em

[edit] to clarify I am using recognize to feed into memories

[31337-4554551n]

Added 2 photos with faces
" 43 faces left to cluster"... Confused

[marcelklehr]

120 faces in total are necessary, so if you already have some people populated then that's not the issue. It says Scheduled background jobs: 0 so that's the issue. You can also run clustering manually by running occ recognize:cluster-faces -b 5000 (batch size to limit memory consumption)

[31337-4554551n]

Thank you, I did that, and am getting

ClusterDebug: Retrieving face detections for user blah
ClusterDebug: Not enough face detections found

for all the users.
Is there a way I can find out why this isn't happening automatically/at all anymore? It did for a bunch to begin with, but I've added more photos since, and I also know it missed a few

[31337-4554551n]

Thank you, please let me know if I can give you any further info

[31337-4554551n]

For more context, I added a bunch of photos today after a trip out.
Today I went for a trip with the girl. Took 8 photos. 7 of her, one with both of us. It's detected 6 of her, and added to her. But one of her is not detected, and the one with both of us isn't detected under either of us (all show up under memories timeline-today)

And the results as above, doesn't think there's any more work to do.

Would a recrawl do anything? Or if it didn't find it the first time, it wouldn't again. Could it break clustering or correctly recognized faces? Or is it completely unnecesary since it's telling me it's aware of 42 faces it just doesn't want to cluster them.

[31337-4554551n]

Is there a way I can find out why this isn't happening automatically/at all anymore? It did for a bunch to begin with, but I've added more photos since, and I also know it missed a few

Yeah, that's strange. It seems like a bug. I'll look into this.

Hi, I was wondering if you've had a chance to look into this. Confirming it's still saying

Face clustering: 42 faces left to cluster, Last clustering run: 8 hours ago, Scheduled background jobs: 0,
A minimum of 120 faces per user is necessary for clustering to kick in

And trying

occ recognize:cluster-faces -b 5000

Still gives me

ClusterDebug: Retrieving face detections for user blah
ClusterDebug: Not enough face detections found

Though I've been using it this whole time, and more photos have been getting identified. But whatever those 42 are appear to be stuck

[marcelklehr]

Those 42 could be from a separate user. The 120 faces is per user. -- Can you perhaps open an issue for this instead of this discussion, then I'll be more probable to follow up 🙈 sorry.

[31337-4554551n]

Thank you again for your time. Issue opened

[31337-4554551n]

Is there perhaps more info or logs I can provide for you to help you troubleshoot the issue?
Or anything I can do to force it to work, perhaps a recrawl? Or would that not be productive.

[marcelklehr]

A recrawl could help. Note, that it's normal that the face detection model does not find all faces, especially if you run in WASM mode.

[31337-4554551n]

I'm also noticing some new photos are getting the proper treatment, and others aren't. Not seeing a pattern