OmniVTON uses pretrained diffusion models with no training to transfer garments between people across shop and street scenes, and extends to multi-human try-on.
Realtime multi-person 2d pose estimation using part affinity fields
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
citation-role summary
background 1
citation-polarity summary
fields
cs.CV 1years
2025 1verdicts
CONDITIONAL 1roles
background 1polarities
unclear 1representative citing papers
citing papers explorer
-
OmniVTON: Training-Free Universal Virtual Try-On
OmniVTON uses pretrained diffusion models with no training to transfer garments between people across shop and street scenes, and extends to multi-human try-on.