Improving performance and explainability of Zero-Shot CLIP | Towards Data Science

Part 2 – Visual classification via Description from LLMs