computer-vision

Zero-shot classification and retrieval via a vision-language model

taught

Classifying an image with no task-specific training by comparing its embedding's cosine similarity against a set of text-prompt embeddings, and using the same shared embedding space for image-text retrieval in either direction.

Taught in

Zero-shot classification and retrieval
Computer Vision
Multimodal assistants: captioning and VQA
Computer Vision

Requires

Nothing — this is foundational.

Required by

Vision-language models recommended

Local knowledge map

Click to expand

Solid arrows are required prerequisites; dashed are recommended or helpful. Lower rows are more foundational. Click any box to jump to that concept.