Face Age Verification Vulnerabilities Under Simple Appearance Manipulations
2026-07-27 • Computer Vision and Pattern Recognition
Computer Vision and Pattern Recognition
AI summaryⓘ
The authors studied how well automated age-checking systems work when underage users try to trick them by changing their appearance, like drawing a mustache or wearing lipstick. They tested seven different models on various altered images and found that these changes can often fool the systems, especially when a drawn beard is added. They also discovered that some groups, like Indians and females, are more affected by these tricks than others. Finally, they looked into simple ways to reduce these biases in the models.
age verificationautomated age estimationvision-based modelsvisual manipulationsbias mitigationTrue NegativesFalse Positivesdemographic biasmachine learning robustnessmultimodal models
Authors
Ioannis Sarridis, Ioannis Kompatsiaris, Symeon Papadopoulos
Abstract
Online platforms increasingly rely on automated age estimation systems to enforce minimum-age policies. Focusing on vision-based models designed for this task, concerns arise regarding their robustness to simple appearance changes that underage individuals may use to bypass such systems, such as drawing a mustache or applying lipstick. In this work, we present a systematic study of age verification robustness by simulating visual alterations that can be easily achieved by underage individuals. We evaluate seven models, including vision, vision-language, and multimodal large language models, across three datasets and four manipulation types. Interestingly, under drawn beard stubble, up to 61% of True Negatives are flipped into False Positives. Furthermore, we investigate how different demographics are affected by such manipulations, finding that Indians are more affected by beard stubble manipulations, while females are more affected than males across all manipulations. Finally, we explore how these biases can be mitigated using bias mitigation methodologies in lightweight linear probe settings.