Install the following packages -

pip install timm
pip install open_clip_torch
pip install scipy

To extract CLIP embeddings -

python embed_clip.py --image_folder="PATH_TO_360_ROCK_IMAGES"

To extract embeddings from (torch)vision models -

python embed_vision.py --image_folder="PATH_TO_360_ROCK_IMAGES"

The image_embeddings.zip contains the generated embeddings from all the models in .csv and .mat format.