📌 Feature Request: Smart Similar Photo Selection, Quality-Based Cleanup, and Auto Categorization via Training

Dear PhotoPrism Team,

I’d like to propose a new set of AI-powered features to improve photo organization and memory management:

1. Smart Similar Photo Selection & Cleanup Suggestions
   Allow users to:

Select a reference photo, and automatically highlight similar photos (e.g. burst shots or duplicates).

Use basic quality assessment (e.g. focus/sharpness, brightness, facial expressions) to suggest deleting blurred or lower-quality alternatives, helping to free up storage.

2. Auto Categorization Based on User-Trained Albums
   Enable users to:

Manually assign a set of photos to an album/folder, essentially “teaching” the system what belongs there.

Based on this training, PhotoPrism would then suggest additional photos that might fit the same album based on:

Faces

Locations

Objects

Time proximity

Scene type

This would create a semi-supervised learning workflow for automatic organization that adapts to each user’s way of grouping memories.

These features would greatly enhance user experience by saving time, reducing clutter, and helping users better curate their photo libraries.

Thanks for the great work on this amazing project!