Generating Realistic Forehead-Creases for User Verification via Conditioned Piecewise Polynomial Curves

TL;DR

This paper introduces a geometric and diffusion-based approach to generate realistic forehead-crease images for user verification, improving system robustness across different datasets.

Contribution

It presents a novel trait-specific image synthesis method using B-spline and Bézier curves combined with diffusion models for enhanced forehead-crease verification.

Findings

Improved verification accuracy with synthetic data augmentation

Effective generation of diverse, realistic crease patterns

Enhanced cross-database verification performance

Abstract

We propose a trait-specific image generation method that models forehead creases geometrically using B-spline and B\'ezier curves. This approach ensures the realistic generation of both principal creases and non-prominent crease patterns, effectively constructing detailed and authentic forehead-crease images. These geometrically rendered images serve as visual prompts for a diffusion-based Edge-to-Image translation model, which generates corresponding mated samples. The resulting novel synthetic identities are then used to train a forehead-crease verification network. To enhance intra-subject diversity in the generated samples, we employ two strategies: (a) perturbing the control points 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 crease patterns. By integrating the proposed synthetic dataset with real-world data, our method significantly improves the performance of forehead-crease verification systems under a cross-database verification protocol.