Search papers, labs, and topics across Lattice.
2
0
4
Stop averaging and start attending: CAMFusion's cross-attentive fusion of multiview vision-language embeddings unlocks state-of-the-art 3D semantic segmentation, even in zero-shot scenarios.
