When Morphs Look Better: Face Image Quality as a Training-Free Signal for Morphing Attack Detection

Hugging Face Daily Papers Papers
morphing-attack-detection face-image-quality biometrics diffusion-attacks evaluation security

Summary

This paper evaluates 17 face image quality assessment methods as training-free signals for morphing attack detection across 20 attack sets, finding that DifFIQA(R) performs best on FRLL while many methods paradoxically assign higher quality to morphs than bona fide samples, and results transfer poorly across datasets.

Face image quality assessment (FIQA) estimates the utility of a face image for automated face recognition and has been proposed as a morphing attack detection (MAD) signal that requires no morph-specific training. We evaluate 17 FIQA methods from all current FIQA categories on 20 morphing attack sets from two datasets, FRLL and FEI, including five diffusion-based attack sets. Each quality score is used directly as a MAD decision score. Unlike prior work, which inverted a quality measure when this lowered its error, we fix the score direction of each method in advance from its definition, so that cases in which morphs receive higher quality than bona fide samples remain visible. On FRLL, DifFIQA(R) achieves a mean D-EER of 11.42%, and together with CLIB-FIQA and eDifFIQA(L) clearly outperforms all other FIQA methods. A strong separation of bona fide and morph scores does not imply good detection performance: on the diffusion-based attacks, up to 11 of 17 methods assign higher quality to morphs than to bona fide samples, and the largest reversed separation in the study coincides with the worst detection result of the study. Performance also depends strongly on the attack: GraFIQs(L), the best detector of DCMorph and Arc2Morph on FRLL, performs poorly on the landmark-based and GAN-based attacks. On FEI, the lowest mean D-EER is 39.75%, and a threshold set on the bona fide data of one dataset does not transfer to the other.
Original Article
View Cached Full Text

Cached at: 11/03/26, 11:36 AM

Paper page - When Morphs Look Better: Face Image Quality as a Training-Free Signal for Morphing Attack Detection

Source: https://huggingface.co/papers/2611.00074

Abstract

Face image quality assessment methods vary widely in detecting diffusion-based and other morphing attacks, with performance depending heavily on attack type and dataset, and high quality scores for morphs often indicate poor detection.

Face image quality assessment(FIQA) estimates the utility of a face image for automated face recognition and has been proposed as amorphing attack detection(MAD) signal that requires no morph-specific training. We evaluate 17 FIQA methods from all current FIQA categories on 20 morphing attack sets from two datasets, FRLL and FEI, including five diffusion-based attack sets. Each quality score is used directly as a MAD decision score. Unlike prior work, which inverted a quality measure when this lowered its error, we fix the score direction of each method in advance from its definition, so that cases in which morphs receive higher quality than bona fide samples remain visible. On FRLL,DifFIQA(R) achieves a meanD-EERof 11.42%, and together withCLIB-FIQAandeDifFIQA(L) clearly outperforms all other FIQA methods. A strong separation of bona fide and morph scores does not imply good detection performance: on thediffusion-based attacks, up to 11 of 17 methods assign higher quality to morphs than to bona fide samples, and the largest reversed separation in the study coincides with the worst detection result of the study. Performance also depends strongly on the attack:GraFIQs(L), the best detector ofDCMorphandArc2Morphon FRLL, performs poorly on the landmark-based and GAN-based attacks. On FEI, the lowest meanD-EERis 39.75%, and a threshold set on the bona fide data of one dataset does not transfer to the other.

View arXiv pageView PDFGitHub0Add to collection

Get this paper in your agent:

hf papers read 2611\.00074

Don’t have the latest CLI?curl \-LsSf https://hf\.co/cli/install\.sh \| bash

Models citing this paper0

No model linking this paper

Cite arxiv.org/abs/2611.00074 in a model README.md to link it from this page.

Datasets citing this paper0

No dataset linking this paper

Cite arxiv.org/abs/2611.00074 in a dataset README.md to link it from this page.

Spaces citing this paper0

No Space linking this paper

Cite arxiv.org/abs/2611.00074 in a Space README.md to link it from this page.

Collections including this paper1

Similar Articles

FoMo: Forking Moment in Generative Trajectory as a Perceptual Distance

Hugging Face Daily Papers

This paper introduces a fully automated data generation pipeline for reference-based image quality assessment metrics, using the forking moment in diffusion models as a perceptual distance proxy that aligns with human visual system and outperforms human-annotated datasets.