Feature self-guidance disperses internal features of flow models during batch generation and applies manifold regularization to increase output diversity while preserving condition alignment.
Diverse text-to-image generation via contrastive noise optimization
4 Pith papers cite this work. Polarity classification is still indexing.
4
Pith papers citing it
years
2026 4verdicts
UNVERDICTED 4representative citing papers
ZeNO frames noise optimization as a path-integral control problem solvable from zeroth-order reward evaluations, connecting to implicit Langevin dynamics for reward-tilted distributions.
Coupled initial noises in diffusion models, with designed dependence but unchanged marginal Gaussians, improve generated image diversity on Stable Diffusion variants while preserving quality and alignment.
FEAT enables fashion editing and virtual try-on from arbitrary design sources including non-apparel imagery using disentangled dual injection and orthogonal-guided noise fusion.
citing papers explorer
-
Don't Settle at the Mode! Mitigating Diversity Collapse in Pretrained Flow Models via Feature Self-Guidance
Feature self-guidance disperses internal features of flow models during batch generation and applies manifold regularization to increase output diversity while preserving condition alignment.
-
Gradient-Free Noise Optimization for Reward Alignment in Generative Models
ZeNO frames noise optimization as a path-integral control problem solvable from zeroth-order reward evaluations, connecting to implicit Langevin dynamics for reward-tilted distributions.
-
Couple to Control: Joint Initial Noise Design in Diffusion Models
Coupled initial noises in diffusion models, with designed dependence but unchanged marginal Gaussians, improve generated image diversity on Stable Diffusion variants while preserving quality and alignment.
-
FEAT: Fashion Editing and Try-On from Any Design
FEAT enables fashion editing and virtual try-on from arbitrary design sources including non-apparel imagery using disentangled dual injection and orthogonal-guided noise fusion.