Search papers, labs, and topics across Lattice.
3
0
6
Achieving 100x parameter efficiency, LoopWM redefines how we approach long-horizon world modeling by introducing iterative latent depth as a new scaling axis.
Rethinking IRSTD as a centroid regression problem with single-point supervision achieves competitive detection performance with significantly reduced computational cost, challenging the dominance of pixel-level segmentation approaches.
Overcome the limitations of existing information-theoretic methods by using structural entropy to learn high-order feature correlations, leading to improved feature selection in multi-view multi-label learning.
