A training-free inpainting method that uses classifier gradients to guide masked region content toward a target class, with hybrid sampling for coherence.
In: NeurIPS (2022)
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.CV 1years
2025 1verdicts
CONDITIONAL 1representative citing papers
citing papers explorer
-
GuidPaint: Class-Guided Image Inpainting with Diffusion Models
A training-free inpainting method that uses classifier gradients to guide masked region content toward a target class, with hybrid sampling for coherence.