DDIS generates training-like images from a frozen classifier by steering Stable Diffusion with batch-normalization statistics and an optimized per-class token, improving data-free distillation and pruning.
Plug-in inversion: Model-agnostic inversion for vision with data augmentations
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
-
When Model Knowledge meets Diffusion Model: Diffusion-assisted Data-free Image Synthesis with Alignment of Domain and Class
DDIS generates training-like images from a frozen classifier by steering Stable Diffusion with batch-normalization statistics and an optimized per-class token, improving data-free distillation and pruning.