RMPL: Relation-aware Multi-task Progressive Learning with Stage-wise Training for Multimedia Event Extraction 事件

RMPL: Relation-aware Multi-task Progressive Learning with Stage-wise Training for Multimedia Event Extraction arXiv:2602.13748v2 Announce Type: replace-cross Abstract: Multimedia Event Extraction (MEE) aims to identify events and their arguments from documents that contain both text and images. It requires grounding event semantics across different modalities. Progress in MEE is limited by the lack of annotated training data. M2E2 is the only established benchmark, but it provides annotations o