Positional Segmentor-Guided Counterfactual Fine-Tuning for Spatially Localized Image Synthesis

ArXi:2603.21213v1 Announce Type: cross Counterfactual image generation enables controlled data augmentation, bias mitigation, and disease modeling. However, existing methods guided by external classifiers or regressors are limited to subject-level factors (e.g., age) and fail to produce localized structural changes, often resulting in global artifacts. Pixel-level guidance using segmentation masks has been explored, but requires user-defined counterfactual masks, which are tedious and impractical.