Image zero-order classification model based on cross knowledge and classification method thereof

Abstract

The invention discloses an image zero-order classification model based on cross knowledge. The image zero-order classification model comprises: a biological classification tree module, which is used for constructing a biological classification tree according to all categories in the data set; the visual feature extraction module that is used for converting images in the data set into one-dimensional visual features; the semantic feature extraction module that is used for converting texts or attributes in the data set into one-dimensional semantic features; the cross knowledge learning module that is used for enriching semantic information of categories; the generative adversarial network module that comprises a generator and a discriminator, the generator generates pseudo visual features from the semantic features, and the discriminator is used for discriminating the authenticity and the category of the image. According to the invention, cross knowledge learning is adopted, more related semantic features can be trained, so that features from semantics to vision are embedded in the ZSL, and the semantic features in the cross-modal learning process are enriched; the model and the method are simple and efficient, and high-accuracy classification results are obtained on multiple authoritative data sets.

Description

technical field [0001] The invention relates to the technical field of image classification, in particular to an image zero-order classification model and a classification method based on cross-knowledge. Background technique [0002] Due to the rapid expansion of data scale and the vigorous development of machine learning models, the field of image classification has become more and more attractive. However, collecting sufficient data sets is time-consuming and laborious, and some data sets are not available. How to correctly and efficiently classify some categories in the absence of some data sets has become one of the main challenges in the field of image classification. [0003] To solve the problem of imperfect data sets, the current mainstream solutions in the field first proposed the concept of Zero-Shot Learning (ZSL). It can identify new categories that have not appeared in the training phase during the test phase, that is, it is used to solve the situation where t...

AI Technical Summary

Problems solved by technology

[0005] 1. Cross-modality problem: The cross-modality problem in visual-semantic embedding leads to incomplete representation of semantic features and visual features during embedding, especially in two categories that are very similar and have no difference in embedding space , which makes it difficult for the model to discriminate and greatly reduces the performance of the model;

[0006] 2. Cross-domain problems: seen classes may rarely (or not) intersect with unseen classes, and the visual appearance of the same attribute or textual description may be significantly different in invisible classes, which will lead to It is difficult for the model to accurately distinguish unseen classes when embedding semantic vectors into visual space

Images