Search papers, labs, and topics across Lattice.
1
0
3
2
Achieve robust text-to-motion generation for long, compositional prompts by planning in the latent space of pre-trained behavioral foundation models, rather than relying on heavy end-to-end motion generators.