1 paper · 1 filter
Muhammad Gohar Javed, Chuan Guo, Li Cheng +1
Generating realistic 3D human-human interactions from textual descriptions remains a challenging task. Existing approaches, typically based on diffusion models, often produce resul…