Search results for: 'Set-CLIP: Exploring Aligned Semantic From Low-Alignment Multimodal Data Through A Distribution View'