Face Presentation Attack Detection

Overview

Face Presentation Attack Detection (FacePAD) aims to distinguish genuine live presentations from artifacts or manipulated inputs used to fool biometric systems. Methods range from texture and motion analysis with convolutional networks to Vision Transformers, frequency-based features, and multimodal models that use text, tools, or structured attack knowledge. Researchers evaluate these systems on benchmarks such as OULU-NPU, SiW, and Replay-Attack, including tests that train on one dataset or attack type and evaluate on another. A central challenge is that strong results on familiar data often do not transfer to new sensors, attack methods, demographic groups, or real-time settings.

Latest papers

CardsList

33 papers6 papers in the last four weeks, against 2 the four weeks before. 0.0% of all new papers.

  1. Face Presentation Attack Detection via Content-Adaptive Spatial Operators

    Feb 21, 2026Shujaat KhanLightweight Neural NetworksFace Presentation Attack Detection

  2. Emergent Morphing Attack Detection in Open Multi-modal Large Language Models

    Feb 17, 2026Marija Ivanovska, Vitomir ŠtrucFace Morphing Attack DetectionImage Forgery Detection