Search papers, labs, and topics across Lattice.
2
0
5
3
VLMs struggle with precise localization in medical imaging, but injecting medical knowledge and using global-local attention can boost grounding accuracy by up to 3% AP50 and 2.6% mIoU.
Achieve state-of-the-art tooth segmentation by teaching multiple student networks on unlabeled multi-source data, each supervised by its own teacher, and enforcing stricter confidence constraints.
