Yongteng Du

Instance-Aligned Semantic Reconstruction for Incomplete Multi-View Clustering

Incomplete multi-view clustering (IMVC) aims to exploit complementary information from multiple views with missing observations. Recent diffusion-based approaches have shown promise for view completion; however, they often fail to capture instance-aligned global correlations across views and suffer