DivControl factorizes ControlNet weights via SVD into shared 'learngenes' and condition-specific 'tailors', routed by a text-conditioned gate, enabling unified control and efficient adaptation to new conditions.
Cock- tail: Mixing multi-modality control for text-conditional im- age generation
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
unclear 1representative citing papers
citing papers explorer
-
DivControl: Knowledge Diversion for Controllable Image Generation
DivControl factorizes ControlNet weights via SVD into shared 'learngenes' and condition-specific 'tailors', routed by a text-conditioned gate, enabling unified control and efficient adaptation to new conditions.