Мультимодальное представление

AbstractThis paper studies the problem of generatinglikely queries for multimodal documents withimages. Our application scenario is enablingefficient “first-stage retrieval” of relevant doc-uments, by attaching generated queries to doc-uments before indexing. …...

AbstractIn contrast to recent advances focusing on high-level representation learning across modalities, in this work we present a self-supervised learning framework that is able to learn a representation that captures …...