ConceptMultimodal Embedding
Multimodal Embedding
A representation learning method that maps data from different modalities, such as images and text, into a unified vector space, so that semantically similar cross-modal samples lie closer together in the embedding space. It is widely used in cross-modal retrieval, classification and related tasks.
Timeline (last 90 days)
Oct 12
Multimodal embedding requires data of different modalities (text, images, audio, etc.) to be mapped into the same coordinate system, so that no cross-model score alignment is needed at retrieval time.
Unverified50%