Abstract
We propose a trait-specific image generation method that models foreheadcreases geometrically using B-spline and B\'ezier curves. This approach ensuresthe realistic generation of both principal creases and non-prominent creasepatterns, effectively constructing detailed and authentic forehead-creaseimages. These geometrically rendered images serve as visual prompts for adiffusion-based Edge-to-Image translation model, which generates correspondingmated samples. The resulting novel synthetic identities are then used to traina forehead-crease verification network. To enhance intra-subject diversity inthe generated samples, we employ two strategies: (a) perturbing the controlpoints of B-splines under defined constraints to maintain label consistency,and (b) applying image-level augmentations to the geometric visual prompts,such as dropout and elastic transformations, specifically tailored to creasepatterns. By integrating the proposed synthetic dataset with real-world data,our method significantly improves the performance of forehead-creaseverification systems under a cross-database verification protocol.