Changelogs:
- The entire code has been updated. Now you have the option to choose from over 500 feature extraction models, including ViT (Vision Transformers).
- Annoy has been removed due to indexing restrictions.
- FAISS is now being used for faster indexing and reindexing of new images.
- Various major issues have been fixed.
- The documentation has been updated.
- GPU support has been added.