Cultural relic image retrieval method and device based on multi-modal representation learning model

The invention provides a cultural relic image retrieval method and device based on a multi-modal representation learning model, and belongs to the technical field of cultural relic image retrieval, and the method comprises the steps: obtaining cultural relic image information and cultural relic text information; the cultural relic character information is used for supplementing or revising attributes of the cultural relic image information; historical relic image feature extraction is carried out on the historical relic image information to obtain historical relic image feature vectors; performing cultural relic character feature extraction on the cultural relic character information to obtain a cultural relic character feature vector; inputting the cultural relic image feature vector and the cultural relic character feature vector into a multi-modal representation learning model to obtain a cultural relic joint feature vector set output by the multi-modal representation learning model; and based on the cultu