Search papers, labs, and topics across Lattice.
4
0
6
2
Small moving instances can now be segmented more effectively than ever, even in low-light and fast-motion scenarios.
Finally, a framework that disentangles the messy process of manga creation, giving you knobs to tweak layout, characters, and text across entire stories.
Achieve near-lossless 60% attention latency reduction in video editing by exploiting query sharpness to dynamically route attention.
VLMs have hidden modularity: selectively knocking out specific layers *improves* performance on downstream tasks by up to 16%, without any training.
