Search papers, labs, and topics across Lattice.
University of Manchester
1
0
3
8
VLMs still struggle to consistently extract structured cultural metadata from images, revealing a critical gap in their ability to reason beyond visual perception across diverse cultural contexts.
