Prompt Me Up: Unleashing the Power of Alignments for Multimodal Entity and Relation Extraction
This work addresses the challenge of multimodal entity and relation extraction for applications like information retrieval, offering incremental improvements through novel alignment techniques.
The paper tackled the problem of extracting entities and relations from text by leveraging multimodal signals from images and text, proposing pre-training objectives for alignment that resulted in an average 3.41% F1 improvement over prior state-of-the-art methods.
How can we better extract entities and relations from text? Using multimodal extraction with images and text obtains more signals for entities and relations, and aligns them through graphs or hierarchical fusion, aiding in extraction. Despite attempts at various fusions, previous works have overlooked many unlabeled image-caption pairs, such as NewsCLIPing. This paper proposes innovative pre-training objectives for entity-object and relation-image alignment, extracting objects from images and aligning them with entity and relation prompts for soft pseudo-labels. These labels are used as self-supervised signals for pre-training, enhancing the ability to extract entities and relations. Experiments on three datasets show an average 3.41% F1 improvement over prior SOTA. Additionally, our method is orthogonal to previous multimodal fusions, and using it on prior SOTA fusions further improves 5.47% F1.