arXiv cs.CV
· Papers
PolypSteer: Counterfactual Endoscopic Synthesis via Training-Free Activation Steering
arXiv:2603.07066v2 Announce Type: replace Abstract: Generative diffusion models are increasingly used for medical imaging data augmentation, but text prompting cannot produce causal training data. Re-prompting rerolls the entire generation trajectory, altering anatomy, texture, and background. Inversion-based editing m