Search papers, labs, and topics across Lattice.
University of Tübingen
1
0
3
Text-to-3D generation of human-object interactions jumps 4-15x in text consistency thanks to a new pipeline trained on data curated with multimodal LLMs.
