Attention-map KL losses plus a PAC-Bayes-style regularizer give small CLIP similarity gains for compositional text-to-image generation, but the theoretical derivation is invalid and the evaluation is under-powered.
Attend-and-excite: Attention-based se- mantic guidance for text-to-image diffusion models
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.CV 1years
2024 1verdicts
REJECT 1representative citing papers
citing papers explorer
-
Unlocking the Potential of Text-to-Image Diffusion with PAC-Bayesian Theory
Attention-map KL losses plus a PAC-Bayes-style regularizer give small CLIP similarity gains for compositional text-to-image generation, but the theoretical derivation is invalid and the evaluation is under-powered.