Small sample multi-element scene image recognition method based on pre-training cross-modal retrieval

• Main Authors: QU XIANGYAN, LEE JIN, GOU GAOPENG, YU JING, XIONG GANG
• Format: Patent
• Language: Chinese; English
• Published: 26.01.2024
• Subjects: CALCULATING; COMPUTING; COUNTING; ELECTRIC DIGITAL DATA PROCESSING; PHYSICS

The invention discloses a small sample multi-element scene image recognition method based on pre-training cross-modal retrieval, and the method comprises the steps: 1), building an image database which comprises images of a plurality of general scene categories and images of a plurality of target scene categories, each scene category comprises a plurality of images, and each image is provided with a corresponding scene category label; 2) extracting features of images of each scene category in the image database by using the scene recognition model to obtain a feature database corresponding to each scene category; generating a feature vector according to the features of the same image in the feature database; 3) for an image A to be subjected to category recognition, extracting the features of the image A by using the scene recognition model and generating a feature vector; 4) the scene recognition model performs similarity calculation on the feature vector of the image A and the feature vectors in the feature