Explicit Over Implicit: Enhancing CNNs Via Complex Structure Tensor Representations for Periocular Recognition

arXiv:2607.15410v1 Announce Type: new Abstract: Our study provides evidence that CNNs struggle to extract orientation features effectively. We show that using the Complex Structure Tensor, which contains compact orientation features with certainties, as input to CNNs consistently improves identification accuracy compared to grayscale inputs alone. Experiments also demonstrated that our inputs, provided by mini-complex convnets, combined with reduced CNN sizes, outperformed full-fledged, prevaili...