CLIP is a powerful Image Model that is trained under the supervision of Natural Language.

We Explore different Semantic Search Capabilities of CLIP.

1. Image - Image Search
2. Text - Image Search
3. Image - Text Search You can find a detailed writeup here

$ pip install -r requirements.txt

To run the Image-Image Search

$ python image_image_search.py

Results:

To run the Image-Text Search

$ python image_text_search.py

Results:

To run the Text-Image Search

$ python text_image_search.py

Results: