Field-Localized Forgery Detection for Digital Identity Documents

2026-05-09

AI Summary

This work addresses the vulnerability of digital identity documents in remote verification scenarios, where localized manipulations—such as face swapping or text tampering—often evade detection by general-purpose forgery detectors. To tackle this challenge, the authors propose FLiD, a novel framework introducing field-level local forgery detection tailored for digital ID documents. FLiD leverages a fine-tuned object detector to localize critical regions and employs a frozen MobileNetV3-Small backbone for feature extraction, followed by an ultra-lightweight classification head with only 191K trainable parameters. Experimental results demonstrate that FLiD achieves AUC scores of 0.880, 0.954, and 0.923 under face, text, and combined attacks, respectively, significantly outperforming baseline methods in terms of Equal Error Rate (EER). Moreover, the model reduces parameter count by 13× and FLOPs by 21× compared to existing approaches.

Abstract

Digital identity verification systems used in remote onboarding rely on document images to authenticate users, making them vulnerable to localized manipulations of key identity fields such as facial photographs and textual information. Existing forgery detection methods, developed primarily for natural-image forensics, show limited transferability to structured identity documents. We propose FLiD, a lightweight field-localized framework that targets critical identity regions rather than processing full-document images. A fine-tuned object detector first localizes face and text fields; a frozen MobileNetV3-Small backbone then extracts compact field-level embeddings, which are classified by lightweight neural network with only 191K trainable parameters. FLiD achieves AUC scores of 0.880 (face), 0.954 (text), and 0.923 (both-field attacks), with corresponding EERs of 18.05%, 11.61%, and 15.16%, representing absolute reductions of 29-35 percentage points over a full-document baseline trained from scratch. FLiD also consistently outperforms general-purpose manipulation detectors (TruFor, MMFusion, UniVAD) across all attack scenarios while requiring 13x fewer parameters and 21x fewer FLOPs

Problem

• digital identity documents
• localized forgery detection
• field-level manipulation
• identity verification
• document forensics

Innovation

• field-localized forgery detection
• digital identity documents
• lightweight framework
• MobileNetV3-Small
• manipulation detection