Abstract: We propose a novel multimodal topic modeling framework to extract and explain latent themes in extensive collections of digitized artworks. Our approach leverages CLIP's contrastive ...
Some results have been hidden because they may be inaccessible to you
Show inaccessible results
Feedback