OrthoMAD: Morphing Attack Detection Through Orthogonal Identity Disentanglement

TL;DR

OrthoMAD introduces a novel regularisation technique for morphing attack detection that disentangles identity information into orthogonal latent vectors, achieving state-of-the-art results across multiple datasets.

Contribution

The paper proposes a new regularisation term promoting orthogonal identity disentanglement in morphing attack detection models, improving detection performance.

Findings

Achieves state-of-the-art results on FRLL dataset.

Performs well across five different morphing types.

Demonstrates robustness when trained on various datasets.

Abstract

Morphing attacks are one of the many threats that are constantly affecting deep face recognition systems. It consists of selecting two faces from different individuals and fusing them into a final image that contains the identity information of both. In this work, we propose a novel regularisation term that takes into account the existent identity information in both and promotes the creation of two orthogonal latent vectors. We evaluate our proposed method (OrthoMAD) in five different types of morphing in the FRLL dataset and evaluate the performance of our model when trained on five distinct datasets. With a small ResNet-18 as the backbone, we achieve state-of-the-art results in the majority of the experiments, and competitive results in the others. The code of this paper will be publicly available.