Guardado en:
| Autores principales: | , , , , , |
|---|---|
| Formato: | Preprint |
| Publicado: |
2026
|
| Materias: | |
| Acceso en línea: | https://arxiv.org/abs/2601.12895 |
| Etiquetas: |
Agregar Etiqueta
Sin Etiquetas, Sea el primero en etiquetar este registro!
|
| _version_ | 1866917209941475328 |
|---|---|
| author | Naseeb, Chan Cheema, Adeel Ashraf Sami, Hassan Afzal, Tayyab Omair, Muhammad Habib, Usman |
| author_facet | Naseeb, Chan Cheema, Adeel Ashraf Sami, Hassan Afzal, Tayyab Omair, Muhammad Habib, Usman |
| contents | The proliferation of sophisticated generative AI models has significantly escalated the threat of synthetic manipulations in identity documents, particularly through face swapping and text inpainting attacks. This paper presents TwoHead-SwinFPN, a unified deep learning architecture that simultaneously performs binary classification and precise localization of manipulated regions in ID documents. Our approach integrates a Swin Transformer backbone with Feature Pyramid Network (FPN) and UNet-style decoder, enhanced with Convolutional Block Attention Module (CBAM) for improved feature representation. The model employs a dual-head architecture for joint optimization of detection and segmentation tasks, utilizing uncertainty-weighted multi-task learning. Extensive experiments on the FantasyIDiap dataset demonstrate superior performance with 84.31\% accuracy, 90.78\% AUC for classification, and 57.24\% mean Dice score for localization. The proposed method achieves an F1-score of 88.61\% for binary classification while maintaining computational efficiency suitable for real-world deployment through FastAPI implementation. Our comprehensive evaluation includes ablation studies, cross-device generalization analysis, and detailed performance assessment across 10 languages and 3 acquisition devices. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2601_12895 |
| institution | arXiv |
| publishDate | 2026 |
| record_format | arxiv |
| spellingShingle | TwoHead-SwinFPN: A Unified DL Architecture for Synthetic Manipulation, Detection and Localization in Identity Documents Naseeb, Chan Cheema, Adeel Ashraf Sami, Hassan Afzal, Tayyab Omair, Muhammad Habib, Usman Computer Vision and Pattern Recognition Machine Learning The proliferation of sophisticated generative AI models has significantly escalated the threat of synthetic manipulations in identity documents, particularly through face swapping and text inpainting attacks. This paper presents TwoHead-SwinFPN, a unified deep learning architecture that simultaneously performs binary classification and precise localization of manipulated regions in ID documents. Our approach integrates a Swin Transformer backbone with Feature Pyramid Network (FPN) and UNet-style decoder, enhanced with Convolutional Block Attention Module (CBAM) for improved feature representation. The model employs a dual-head architecture for joint optimization of detection and segmentation tasks, utilizing uncertainty-weighted multi-task learning. Extensive experiments on the FantasyIDiap dataset demonstrate superior performance with 84.31\% accuracy, 90.78\% AUC for classification, and 57.24\% mean Dice score for localization. The proposed method achieves an F1-score of 88.61\% for binary classification while maintaining computational efficiency suitable for real-world deployment through FastAPI implementation. Our comprehensive evaluation includes ablation studies, cross-device generalization analysis, and detailed performance assessment across 10 languages and 3 acquisition devices. |
| title | TwoHead-SwinFPN: A Unified DL Architecture for Synthetic Manipulation, Detection and Localization in Identity Documents |
| topic | Computer Vision and Pattern Recognition Machine Learning |
| url | https://arxiv.org/abs/2601.12895 |