Adding InfoNCE loss and a SegFormer mask encoder to a Giambi-style face diffusion pipeline lowers the reported FID from 74.07 to 70.98, and to 63.85 when segmentation is included.
Score-Based Generative Modeling through Stochastic Differential Equations,
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
-
Enhancing Diffusion Face Generation with Contrastive Embeddings and SegFormer Guidance
Adding InfoNCE loss and a SegFormer mask encoder to a Giambi-style face diffusion pipeline lowers the reported FID from 74.07 to 70.98, and to 63.85 when segmentation is included.