Evaluating zero-shot classification models with this dataset

#3
by fhvilshoj - opened

Hi!
Thanks for your efforts in collecting this dataset. It's helping push forward the field of AI!

I wanted to show how the efforts of collecting this dataset is contributing to downstream evaluations of open-source zero-shot classification models like CLIP and domain-specific ones like BioCLIP.

Have a look here: https://github.com/encord-team/text-to-image-eval

medical-linear-probe-accuracy.png

Would be curious to hear how this stacks up against your findings!

Sign up or log in to comment