| Abstrakt: |
Label classification aims to select the most relevant subset of labels from a set of labels to tag an instance, which has become a hot issue in the field of artificial intelligence. Traditional multilabel learning methods mainly focus on identifying single-modal data, with limited research on mining high-order correlation between multi-modal data. To address the issue of insufficient representation of high-order correlations between multi-modal data in multi-label scenarios, this paper proposed a multimodal multi-label classification method based on hypergraphs. The hypergraph model is introduced to model the high-order correlations of multi-modal data, and the fusion of multi-modal features and hyperedge convolution operation are utilized to achieve the mining of multi-modal data relationships and feature recognition, thus improving the performance of multi-modal multi-label classification. Experiments were conducted on the movie genre classification task, and the proposed method was compared with traditional methods. The experimental results show that the proposed method outperforms the comparison methods in terms of accuracy, precision, and F1 score, demonstrating the effectiveness of the method. [ABSTRACT FROM AUTHOR] |
