GitHub - slavabarkov/tidy: Offline semantic Text-to-Image and Image-to-Image search on Android powered by quantized state-of-the-art vision-language pretrained CLIP model and ONNX Runtime inference engine
github.com
Offline semantic Text-to-Image and Image-to-Image search on Android powered by quantized state-of-the-art vision-language pretrained CLIP model and ONNX Runtime inference engine - slavabarkov/tidy

cross-posted from: https://lemm.ee/post/43866352

Tidy- Offline semantic Text-to-Image and Image-to-Image search on Android powered by quantized state-of-the-art vision-language pretrained CLIP model and ONNX Runtime inference engine

Features

  • Text-to-Image search: Find photos using natural language descriptions.
  • Image-to-Image search: Discover visually similar images.
  • Automatic indexing: New photos are automatically added to the index.
  • Fast and efficient: Get search results quickly.
  • Privacy-focused: Your photos never leave your device.
  • No internet required: Works perfectly offline.
  • Powered by OpenAI’s CLIP model: Uses advanced AI for accurate results.

F-Droid

  • OtterAEnglish
    3·
    2 months ago

    Oh interesting

    Last update was 12 months ago, are there any newer models out now that it could use?

    edit:

    Wow this is decent. I didn’t have too many photos on my phone, but it’s able to identify some basic animals and plants