A single-step rectified-flow diffusion model trained on synthetic Blender renders generates controllable shadows for object images, with a new public benchmark.
Object- drop: Bootstrapping counterfactuals for photorealistic object removal and insertion, 2024
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
background 1representative citing papers
citing papers explorer
-
Controllable Shadow Generation with Single-Step Diffusion Models from Synthetic Data
A single-step rectified-flow diffusion model trained on synthetic Blender renders generates controllable shadows for object images, with a new public benchmark.