Search papers, labs, and topics across Lattice.
3
0
5
Treating raw visual images as action representations, iMac significantly boosts prediction accuracy and task success in robotic manipulation, outperforming traditional action vector methods.
By allowing non-keyframes to skip denoising steps, RhymeFlow achieves faster video generation without compromising on visual quality.
$\tau_0$-WM outperforms traditional models by seamlessly integrating action prediction and evaluation, leading to superior performance in complex robotic tasks.
