Hierarchical Multi-modal Transformer for Cross-modal Long Document Classification
This addresses the challenge of effectively utilizing multi-modal data in long documents for classification, which is an incremental improvement over prior work focused on short sequences.
The paper tackles the problem of classifying long documents with hierarchical text and embedded images by proposing a Hierarchical Multi-modal Transformer (HMT) that conducts multi-modal feature interaction and fusion, and it outperforms state-of-the-art methods on four datasets.
Long Document Classification (LDC) has gained significant attention recently. However, multi-modal data in long documents such as texts and images are not being effectively utilized. Prior studies in this area have attempted to integrate texts and images in document-related tasks, but they have only focused on short text sequences and images of pages. How to classify long documents with hierarchical structure texts and embedding images is a new problem and faces multi-modal representation difficulties. In this paper, we propose a novel approach called Hierarchical Multi-modal Transformer (HMT) for cross-modal long document classification. The HMT conducts multi-modal feature interaction and fusion between images and texts in a hierarchical manner. Our approach uses a multi-modal transformer and a dynamic multi-scale multi-modal transformer to model the complex relationships between image features, and the section and sentence features. Furthermore, we introduce a new interaction strategy called the dynamic mask transfer module to integrate these two transformers by propagating features between them. To validate our approach, we conduct cross-modal LDC experiments on two newly created and two publicly available multi-modal long document datasets, and the results show that the proposed HMT outperforms state-of-the-art single-modality and multi-modality methods.