Semantic segmentation method and device based on few samples, electronic equipment and storage medium

Abstract

The invention provides a semantic segmentation method and device based on few samples, electronic equipment and a computer readable storage medium, and the method comprises the steps: enabling a target image combination to serve as the input of a trained feature extraction network, and obtaining the image features of each image in the target image combination; calculating an initial class prototype corresponding to the to-be-identified target class based on the image feature, the foreground mask and the background mask of each support image; determining a temporary class prototype of the queryimage based on the image features of a query image and the initial class prototype; fusing the initial class prototype and the temporary class prototype to obtain a final state class prototype; for the image features of the query image, calculating cosine similarity with the final state class prototype pixel by pixel to obtain a first cosine similarity graph; and determining a prediction foreground mask corresponding to the to-be-identified target category according to the first cosine similarity graph. According to the scheme, under the condition that the number of the sample images corresponding to the to-be-recognized target category is small, accurate semantic segmentation can be achieved with the help of a small number of supporting images.

Description

technical field [0001] The present application relates to the technical field of retail management, in particular to a method and device for semantic segmentation based on few samples, electronic equipment, and a computer-readable storage medium. Background technique [0002] The proportion of retail goods refers to the proportion of goods in the distribution channel. Generally, the area ratio of the commodity on the image of the distribution scene can be used as the retail commodity ratio of the commodity. For example, the ratio of the display area of ​​a certain brand of ice cream in a freezer filled with ice cream to the total area of ​​the freezer can be considered as the proportion of retail products of this brand of ice cream. [0003] Semantic segmentation (semantic segmentation) is a common technology for calculating the proportion of commodities. Semantic segmentation is a pixel-level classification, which refers to dividing each pixel in the picture into a specifi...

AI Technical Summary

Problems solved by technology

This results in fewer sample images corresponding to such commodities, which cannot meet the training requirements, making the segmentation accuracy of the semantic segmentation network insufficient

Images