Disentangled Fine-Grained Prototype Learning for Incomplete Image-Tabular Classification

Arxiv CS.CV (computer vision)

Proposes a prototype learning method to handle missing modalities in image-tabular multimodal classification. Addresses challenges in applications like product understanding and medical diagnosis.

multimodal

Arxiv CS.CV (computer vision)·Feixiang ZhouJianyang XieZhuangzhi GaoQinkai YuFu WangYuheng FanJing LiZheheng JiangYitian ZhaoYanda MengHe ZhaoGregory Y. H. LipYalin Zheng

6 days ago