Setting cross-attention maps to a fixed uniform distribution over text tokens reduces reconstruction error in DDIM inversion and improves text-guided editing when combined with adaptive mask blending.
Continual vision-language retrieval via dynamic knowledge rectification
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
2024 1verdicts
CONDITIONAL 1roles
background 1polarities
unclear 1representative citing papers
citing papers explorer
-
Uniform Attention Maps: Boosting Image Fidelity in Reconstruction and Editing
Setting cross-attention maps to a fixed uniform distribution over text tokens reduces reconstruction error in DDIM inversion and improves text-guided editing when combined with adaptive mask blending.