Maybe do some deduplication first? https://github.com/qarmin/czkawka
this post was submitted on 22 Jun 2026
34 points (100.0% liked)
Selfhosted
60644 readers
643 users here now
A place to share alternatives to popular online services that can be self-hosted without giving up privacy or locking you into a service you don't control.
Rules:
-
Be civil.
-
No spam.
-
Posts are to be related to self-hosting.
-
Don't duplicate the full text of your blog or readme if you're providing a link.
-
Submission headline should match the article title.
-
No trolling.
-
Promotion posts require active participation, with an account that is at least 30 days old. F/LOSS without a paywall has exceptions, with requirements. See the rules link for details. Tags [CBH] or [AIP] are required, see the links in Rule 8 for details.
-
AI-related discussions and AI-involved promotional posts have additional requirements for tagging, as noted in Rule 7 and the AI & Promotional Post Expanded Rules post, and find example disclosures here.
Resources:
- selfh.st Newsletter and index of selfhosted software and apps
- awesome-selfhosted software
- awesome-sysadmin resources
- Self-Hosted Podcast from Jupiter Broadcasting
Any issues on the community? Report it using the report flag.
Questions? DM the mods!
founded 3 years ago
MODERATORS
czkawka
That certainly would thin the pile and cut down on the manual labor aspect. It uses a hashing method, iirc. Used it to thin out duplicate audio files, being careful not to delete files that might have the same filename but one would be a live rendition, and a studio/album rendition of the same song. Jimi Hendrix, in my experience, is notorious for this. One of the things that I dig about him, is that he never really performed the same song the same way. He sort of just really went with a stream of consciousness and pulled it off quite well.
I am unsure how you would go about that without AI. You could probably write a python script that hooks in with AI, that ranks by focus, brightness, saturation or other such criteria. However, I suspect that there would have to be a fair amount of manual labor to do that. That's an interesting request. I'll watch the thread and see what the outcome is.
What are the criteria for "best"? If you can define them sufficiently, a self hosted llm, with vision (I'm thinking a Qwen3.6-35B or even a straight Qwen3.4 7B VL) might fit the bill. It's a classification job - LLMs are good at those.
Does immich have dedupe / quality eval plug ins? They might do the job too.
You could write something (or get Claude to code you something) but option 1 seems simplest.
The problem I see here is that the best images are those that have been edited, i.e. rotated, cropped, and adjusted brightness/contrast/saturation/white balance. Like I can have a thousand snaps from my camera, but they'll only look good once I pick them out and edit them.
I suppose, if you have a consistent camera with photos in consistent conditions, you could apply edits in bulk. And I know Google has automated crop/rotate/etc features in the Google Photos app, so maybe you could find a self-hosted tool smart enough to do that across all your photos, then make a review pass to pick out the good ones.