Object relationship prediction method and device, terminal equipment and readable storage medium

Abstract

The invention is suitable for the technical field of image recognition, and provides an object relationship prediction method and device, terminal equipment and a readable storage medium, and the method comprises the steps: obtaining the feature information of a plurality of objects in a to-be-detected image and the visual feature vector of the to-be-detected image through a preset object detector; obtaining non-visual feature vectors of the plurality of objects according to the feature information of each object; and obtaining a predicate prediction result between the two objects through a preset predicate prediction model according to the non-visual feature vector and the visual feature vector between the two objects. When predicates are predicted, the position vector between the two objects and the semantic embedding vector of the two objects are considered; the relation between the zero sample object and the existing object can be determined according to the semantic embedding vector of the zero sample object and the position of the zero sample object when the object label of the zero sample appears, so that the prediction accuracy of the relation between the two objects when the zero sample appears is effectively improved.

Description

technical field [0001] The present application belongs to the technical field of image recognition, and in particular relates to an object relationship prediction method, device, terminal equipment and readable storage medium. Background technique [0002] With the development of technology, great progress has been made in the recognition and detection of individual objects in image recognition, however, the detection of the relationship between objects still needs to be improved. [0003] In the prior art, the relationship between different objects is regarded as a subject-verb-object triplet, the subject and object are different objects, and the predicate is the relationship between these two objects. Through Visual Genome (VG) The data set predicts the predicate category between two objects, wherein the VG dataset includes a fixed number of object categories and predicate categories that match the object category. [0004] However, since the number of object categories a...

AI Technical Summary

Problems solved by technology

[0005] The embodiment of the present application provides an object relationship prediction method, device, terminal equipment, and readable storage medium, which can improve the problem that the relationship between two objects cannot be accurately predicted and the obtained prediction results are not accurate enough when there are zero samples.

Images