PCMEA introduces a semi-supervised approach for multi-modal entity alignment, enhancing alignment quality through pseudo-label calibration and contrastive learning.
Proposing a novel approach using Dirichlet energy to achieve semantic consistency in multi-modal entity alignment.
Proposing the DESAlign framework to address semantic inconsistency in multi-modal entity alignment.
Pseudo-Label Calibration verbessert Multi-Modal Entity Alignment in semi-überwachten Einstellungen.
The proposed MIMEA framework effectively realizes multi-granular interaction mechanisms within and across modalities to enhance multi-modal knowledge representation and alignment.