Đăng nhập
thông tin chi tiết - Multimodal Alignment Framework
Entity-Aware Multimodal Alignment Framework for News Image Captioning Study
The authors propose an Entity-Aware Multimodal Alignment Framework to improve news image captioning by addressing challenges with entity recognition in MLLMs.
Entity-Aware Multimodal Alignment Framework for News Image Captioning: Improving Entity Recognition in News Image Captions
Verbesserung der Entitätenerkennung in Nachrichtenbildunterschriften durch ein multimodales Ausrichtungsframework.