FAM Diffusion lets pre-trained Stable Diffusion models generate images at up to 4x resolution without training by steering low-frequency structure and attention maps from a native-resolution pass.
Scalecrafter: Tuning-free higher- resolution visual generation with diffusion models
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.CV 1years
2024 1verdicts
CONDITIONAL 1representative citing papers
citing papers explorer
-
FAM Diffusion: Frequency and Attention Modulation for High-Resolution Image Generation with Stable Diffusion
FAM Diffusion lets pre-trained Stable Diffusion models generate images at up to 4x resolution without training by steering low-frequency structure and attention maps from a native-resolution pass.