Ranking-learning-based multi-label zero sample classification method

Abstract

The invention relates to a multi-label image classification technology oriented to the field of multimedia content understanding and analysis so as to construct a new classification model and realizea correlation-level-based multi-label classification algorithm. A ranking-learning-based multi-label zero sample classification method provided by the invention comprises the following steps: at a feature extraction stage, carrying out feature description of different modes by using an existing feature extractor to obtain a training data set; at a multi-mode feature transformation stage, giving atraining sample set pair including an original image and a corresponding label between which determined marking information is provided so as to train the model; and at a classification marking stage,giving an original image of a testing sample and a possible label to carry out testing. At the classification marking stage, the correspondence relationship between the original image and the label is not determined. The ranking-learning-based multi-label zero sample classification method is mainly applied to the multi-label image classification occasions.

Description

technical field [0001] The invention relates to a multi-label image classification technology oriented to the field of multimedia content understanding and analysis, and specifically relates to a multi-label zero-sample classification method based on ranking learning. Background technique [0002] With the rapid development of information technology, a large number of multimedia data such as images and videos have emerged, which has become one of the important ways for people to obtain information. Human cognition of image information is to distinguish and label different types of targets reflected by it. However, due to the sharp increase in the types of things and the continuous refinement of the types, the unequal image and label information make it difficult for traditional image classification techniques to meet actual needs. The emergence of zero-shot learning solves the problem of missing labels to a certain extent. In practical applications, different regions of an ...

AI Technical Summary

Problems solved by technology

There are two major problems in directly analyzing and processing samples: 1) high computational complexity and dimensionality disaster; 2) semantic gap, which has become a key issue that seriously restricts the field of multimedia content analysis and retrieval

The above techniques are mainly used for common single-label classification problems. The generalization ability for multi-label zero-sample labeling is insufficient, and the corresponding labeling result of the sample has only one label, which cannot meet the actual requirements of users.

Images