It was one of the early breakthrough Deep Learning algorithms from OpenAI that combined image and text understanding by explicitly encoding them into the same space. The model was given images and captions and rewarded based on the similarity of their embeddings.
What is a CLIP?
It was one of the early breakthrough Deep Learning algorithms from OpenAI that combined image and text understanding by explicitly encoding them into the same space. The model was given images and captions and rewarded based on the similarity of their embeddings.
https://openai.com/index/clip/