SynRDinBAS: Synthetic Realities and Data in Biometric Analysis and Security

Recent advances in generative AI, including generative and foundation models, have transformed data-driven research in computer vision and biometrics by enabling the creation of increasingly realistic synthetic visual and biometric data. Such data offers important opportunities for addressing data scarcity, improving demographic coverage, supporting privacy-preserving learning, and enabling controlled experimentation and data augmentation. At the same time, increasingly realistic synthetic and manipulated content introduces significant risks for biometric security, identity protection, and digital trust, particularly when it becomes difficult to distinguish from authentic content or is used for impersonation, manipulation, and other malicious purposes.

The Synthetic Realities and Data in Biometric Analysis and Security (SynRDinBAS) Workshop is organized at the IEEE/CVF Winter Conference on Applications of Computer Vision (WACV) 2027.

The workshop aims to explore the generation, application, evaluation, and forensic analysis of synthetic data and realities in computer vision, biometrics, and security. It will address both the opportunities offered by synthetic data and the associated challenges related to biometric security, privacy, fairness, authenticity, and responsible use. 

Topics of interest include, but are not limited to:

• Generative and foundation models for synthetic biometric and visual data

• Controllable, identity-preserving, multimodal, and privacy-preserving generation

• Synthetic data for augmentation, fairness, rare cases, and domain adaptation

• Synthetic-to-real generalization and evaluation of data utility and quality

• Deepfakes, face morphs, virtual identities, avatars, and partially manipulated content

• Attacks on biometric and identity-verification systems using generative AI

• Detection, attribution, localization, and explanation of synthetic content

• Robustness to unseen generators, post-processing, and distribution shifts

• Content provenance, watermarking, authenticity verification, and traceability

• Bias, privacy leakage, ethics, governance, and responsible use of synthetic data

• New datasets, benchmarks, evaluation protocols, and real-world applications

Find details about paper submission HERE.