tedornitier

tedornitier / immich-pet-tagger

Public

Automatic pet tagging for Immich using CLIP embeddings and YOLO

44
1
100% credibility
Found May 14, 2026 at 50 stars -- GitGems finds repos before they trend. Get early access to the next one.
Sign Up Free
AI Analysis
Python
AI Summary

A companion tool for Immich that uses a few reference photos to automatically detect and tag pets in your photo library, adding them to the People section.

How It Works

1
🔧 Set up Pet Tagger

You download and start the pet tagger app right next to your Immich photo library on your computer.

2
🔗 Connect to your photos

You enter a special access code from Immich so the tagger can see and update your photo library.

3
🌐 Open the web page

You visit a simple web interface in your browser to manage everything easily.

4
🐱 Add your pets

You create entries for each pet with a fun name, short description like 'fluffy tabby cat', and optional date ranges when they joined your family.

5
📸 Pick example photos

You search for and select 10-20 clear, varied photos of your pet to teach the tagger what they look like, plus mark similar but wrong photos as 'not mine' to sharpen its eye.

6
🔍 Test and scan your library

You run quick tests on recent photos to check accuracy, refine as needed, then scan your whole collection from your pet's join date.

🎉 Pets in People section

New uploads get your pets auto-tagged every few minutes, and they appear in Immich's People view just like family members, bringing smiles every time you browse.

Sign up to see the full architecture

5 more

Sign Up Free

Star Growth

See how this repo grew from 50 to 44 stars Sign Up Free
Repurpose This Repo

Repurpose is a Pro feature

Generate ready-to-use prompts for X threads, LinkedIn posts, blog posts, YouTube scripts, and more -- with full repo context baked in.

Unlock Repurpose
AI-Generated Review

What is immich-pet-tagger?

This Python Docker sidecar automatically tags your pets in Immich photos, treating cats, dogs, or other animals like people in the People section. Provide a few reference photos via a web UI, add a description like "orange tabby cat," and it uses CLIP embeddings with YOLO detection to classify and tag new uploads every 5 minutes—no cloud, no training hassle. Setup involves an Immich API key and docker-compose up, with pre-built images for NVIDIA, AMD ROCm, or CPU.

Why is it gaining traction?

It stands out with a polished UI for importing Immich-detected "people" as pets, finding reference candidates via smart search, and surfacing "not my pet" negatives to cut false positives—far easier than manual tagging. Features like date-range limits for deceased pets, low-confidence review, and automatic polling make it a set-it-and-forget-it automatic pet feeder for your photo library, complete with live scan stats and thumbnail proxy. Multi-GPU support and incremental scans keep it efficient even on large libraries.

Who should use this?

Immich self-hosters with pet photos who want faces grouped without Photoprism or cloud AI. Pet owners tired of untagged puppy pics in family albums, or homelab tinkerers extending Immich like an automatic pet door for animal recognition. Ideal if you have 10k+ photos and a GPU for fast backfills.

Verdict

Grab it if you're on Immich—solid docs and Docker ease make the 44 stars and 1.0% credibility score forgivable for a fresh solo project; expect iterative tweaks as it matures. Skip for production without testing your setup first.

(198 words)

Sign up to read the full AI review Sign Up Free

Similar repos coming soon.