Search papers, labs, and topics across Lattice.
4
2
7
36
Test-time training with KV binding isn't memorization, it's secretly a learned linear attention mechanism, unlocking architectural simplifications and parallelization.
Synthesizing realistic radar data from camera images is now possible, bridging the gap between visual and radar perception for autonomous driving.
Current NVS evaluation metrics are misleading, so this paper introduces a task-aware framework using Zero123 features that actually aligns with human perception of quality and faithfulness.
Control video generation with unprecedented precision, using only crude user-provided motion cues, thanks to a training-free "dual-clock denoising" approach that aligns motion where you want it, and lets the diffusion model fill in the rest.
