immich-app/immich
1
0
Fork 0
mirror of https://github.com/immich-app/immich.git synced 2026-02-05 00:30:57 +03:00
Code Issues 12 Packages Projects Releases 10 Wiki Activity

[BUG] Object detection pipeline not running #68

New Issue
Closed
opened 2026-02-04 17:08:25 +03:00 by OVERLORD · 4 comments
OVERLORD commented 2026-02-04 17:08:25 +03:00
Owner
Copy Link

Originally created by @Eidenz on GitHub (May 10, 2022).

Originally assigned to: @alextran1502 on GitHub.

I was wondering how you implemented object detection. Does it run periodically, or is it supposed to be called after an image is uploaded?

It's barely mentioned in the readme, and I don't see any tags or calls in microservices logs for now (just started uploading some pics to test).

You mention it's based on COCO SSD, do you consider offering alternatives in the future?

Originally created by @Eidenz on GitHub (May 10, 2022). Originally assigned to: @alextran1502 on GitHub. I was wondering how you implemented object detection. Does it run periodically, or is it supposed to be called after an image is uploaded? It's barely mentioned in the readme, and I don't see any tags or calls in microservices logs for now (just started uploading some pics to test). You mention it's based on COCO SSD, do you consider offering alternatives in the future?
OVERLORD commented 2026-02-04 17:08:44 +03:00
Author
Owner
Copy Link

@alextran1502 commented on GitHub (May 10, 2022):

The object detection mechanism is using the pre-trained model from Google, the weight file is downloaded when the microservices container starts up. Whenever an image is uploaded successfully, an event will be queued up and sent to the microservices container, where it picks up the event in a controlled manner and run through the object detection APIs from Tensorflow

I have plans to let people use their own models and their own machine learning server as well, that is why I split up the machine learning portion into its own application. So, if people have their own model, they can point the server to their endpoint and just give the server the result it is expecting

@alextran1502 commented on GitHub (May 10, 2022): The object detection mechanism is using the pre-trained model from Google, the weight file is downloaded when the microservices container starts up. Whenever an image is uploaded successfully, an event will be queued up and sent to the microservices container, where it picks up the event in a controlled manner and run through the object detection APIs from Tensorflow I have plans to let people use their own models and their own machine learning server as well, that is why I split up the machine learning portion into its own application. So, if people have their own model, they can point the server to their endpoint and just give the server the result it is expecting
OVERLORD commented 2026-02-04 17:08:47 +03:00
Author
Owner
Copy Link

@Eidenz commented on GitHub (May 11, 2022):

Okay I see, thanks for the info.

But then I think my installation may have a problem. Because 0 detections on 200 pictures, and no logged attempt in the microservices..

@Eidenz commented on GitHub (May 11, 2022): Okay I see, thanks for the info. But then I think my installation may have a problem. Because 0 detections on 200 pictures, and no logged attempt in the microservices..
OVERLORD commented 2026-02-04 17:08:50 +03:00
Author
Owner
Copy Link

@alextran1502 commented on GitHub (May 11, 2022):

Thank you for the info, I just tested and this is a bug.

@alextran1502 commented on GitHub (May 11, 2022): Thank you for the info, I just tested and this is a bug.
OVERLORD commented 2026-02-04 17:08:58 +03:00
Author
Owner
Copy Link

@alextran1502 commented on GitHub (May 11, 2022):

The issue has been fixed now, it will be built with the latest tag on the Dockerhub shortly

@alextran1502 commented on GitHub (May 11, 2022): The issue has been fixed now, it will be built with the `latest` tag on the Dockerhub shortly
Sign in to join this conversation.
No Branch/Tag Specified
Branches Tags
Labels
Clear labels
accessibility
changelog:enhancement
changelog:security
changelog:translation
cli
date-time
documentation
external-library
format
good first issue
needs-answer
nice to have
sharing
tech-debt
📱mobile
🖥️web
🗄️server
🧠machine-learning
No Label
Milestone
No items
No Milestone
Projects
Clear projects
No project
Assignees
Clear assignees
OVERLORD
No Assignees
1 Participants
Notifications
Due Date
No due date set.
Dependencies

No dependencies set.

Reference: immich-app/immich#68
Delete Branch "%!s()"

Deleting a branch is permanent. Although the deleted branch may continue to exist for a short time before it actually gets removed, it CANNOT be undone in most cases. Continue?