Flux-Sculptor achieves precise text-driven portrait editing by first localizing the target facial region with a trained mask locator, then applying mask-guided latent fusion in early denoising steps and attention value fusion in later steps.
Title resolution pending
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
-
Flux-Sculptor: Text-Driven Rich-Attribute Portrait Editing through Decomposed Spatial Flow Control
Flux-Sculptor achieves precise text-driven portrait editing by first localizing the target facial region with a trained mask locator, then applying mask-guided latent fusion in early denoising steps and attention value fusion in later steps.