Search papers, labs, and topics across Lattice.
1
0
3
A simple adapter that leverages optimal transport and cross-attention can significantly boost the ability of multimodal LLMs to detect AI-generated images by better fusing artifact and semantic features.
