A per-garment virtual try-on pipeline that trains a GAN from a two-minute real-human video capture and uses a hybrid pose-plus-DensePose input to synthesize the garment with accurate alignment.
Title resolution pending
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.GR 1years
2025 1verdicts
CONDITIONAL 1representative citing papers
citing papers explorer
-
Low-Barrier Dataset Collection with Real Human Body for Interactive Per-Garment Virtual Try-On
A per-garment virtual try-on pipeline that trains a GAN from a two-minute real-human video capture and uses a hybrid pose-plus-DensePose input to synthesize the garment with accurate alignment.