LLM-generated text-to-image prompts derived from author style sheets produce images that ten raters judged as moderately faithful (4.08/5), but the evaluation has no control condition and the dataset link is a placeholder.
Generative Adversarial Nets
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
background 1representative citing papers
citing papers explorer
-
Personalized Image Generation from an Author Writing Style
LLM-generated text-to-image prompts derived from author style sheets produce images that ten raters judged as moderately faithful (4.08/5), but the evaluation has no control condition and the dataset link is a placeholder.