Search papers, labs, and topics across Lattice.
B2∑i,jk(gi,gj)⏟repulsion, exact−
5
1
4
3
A single representation can be easily manipulated, but a balanced ensemble of encoders dramatically enhances image generation quality and robustness.
SurgVista achieves unprecedented visual fidelity and interaction accuracy in surgical simulations, outperforming state-of-the-art models as prediction horizons extend.
Frozen video generators already possess the internal signals needed to significantly improve the physical realism of their outputs, without any additional training.
Modeling dynamic 3D scenes gets a serious upgrade: DeGO's deformable Gaussians and foundation model distillation boost occupancy prediction accuracy by 13.5% on human-centric instances.
Generate minute-long, high-fidelity animations without visual degradation or character drift using a surprisingly simple latent flow restoration technique.
