github imagegenius/docker-immich v1.112.0-ig316

3 months ago

ImageGenius Changes:

support GPU acceleration with CUDA for machine-learning

immich Changes:

v1.112.0

Highlights

Welcome to the release of v1.112.0 of Immich. This release packs with new UI features and enhancements, and many bugs are getting squashed. We hope you enjoy this release. Some of the key highlights include:

  • Material color theming for the mobile app
  • Open in map view on the web
  • Star rating
  • Screen stays on and auto darkens when performing a long session of foreground upload
  • Notable fix: Uploading assets from iCloud causing the iOS mobile app to crash
  • Notable fix: Photos and videos belonging to a person are now all displayed in the person's view on the mobile app

Material color theming for the mobile app

The app has had a makeover to prepare for the new school year. You can now choose a different primary color in the app, and the new color will be applied across the app's elements. There are a variety of minor tweaks in UI elements to make the app appear more consistent and aesthetically pleasing.

Open in map view on the web

You can now navigate to the map directly from the asset's detail panel by clicking on the map icon at the top right corner of the minimap.

openinmap

Star rating

As one of the most commonly used pieces of information for photographers, the star rating information in EXIF will now be included in the extracted data. You can run the extract metadata job for all to populate this information in your instance.

image

Enabling this feature in User Settings > App Settings > Star Rating lets you view and set the star rating information in the detail panel.

The edited information will be saved in the companion sidecar XMP file.

settings

Always-on screen for foreground upload

When entering the backup screen and running foreground backup, the screen will now stay on and automatically darken after 30 seconds.

New search models

Immich now supports a set of newer models with very competitive performance. See #11468 for details and visualizations.

A special mention goes to ViT-B-16-SigLIP__webli, which outperforms several much larger ViT-L models despite being the same size as the default model. However, it is a bit slower than the default. At the same size, ViT-B-16-SigLIP-384__webli is among the best models period, at the cost of being much slower than other models of its size. It's ideal for memory-constrained environments that still want the best quality.

Support Immich

If you find the project helpful, you can support Immich by purchasing a product key at https://buy.immich.app.

Cheers! 🍻

What's Changed

🚀 Features

🌟 Enhancements

🐛 Bug fixes

📚 Documentation

🌐 Translations

New Contributors

Full Changelog: immich-app/immich@v1.111.0...v1.112.0

Don't miss a new docker-immich release

NewReleases is sending notifications on new releases.