3ʳᵈ  International Workshop on Synthetic Data for Face and Gesture Analysis 

3ʳᵈ   International Workshop on 

Synthetic Data for Face and Gesture Analysis (SD-FGA) 

About the workshop 

The SD-FGA workshop is organized within the scope of IEEE FG 2026 and all accepted and presented papers will be published within the main "IEEE FG 202 Proceedings" that will appear on IEEE Xplore. The workshop will be held in Kyoto, Japan,  on 26 May 2026 in the afternoon. The workshop is co-organized by the ARIS funded project "DeepFake DAD".

Motivation

The landscape of computer vision and artificial intelligence continues to evolve rapidly with the rise of powerful generative models, fundamentally transforming how data-driven challenges are addressed. Generative models ranging from early methods to cutting-edge architectures such as GANs, VAEs, and diffusion models now enable the creation of synthetic data with unprecedented realism and diversity. This progress is particularly transformative in domains such as facial analysis, gesture recognition, and behavioral understanding, where privacy, ethics, and data diversity remain critical concerns.

Building on the success of the first and second editions of the Synthetic Data for Face and Gesture Analysis (SD-FGA) workshop, we would like to organize the third edition of the SD-FGA workshop as part of FG 2026. This third edition aims to expand the discussion toward synthetic data for attack generation and detection, encompassing both physical and behavioral aspects. As AI systems increasingly operate in open-world environments, understanding and mitigating adversarial and spoofing attacks, whether through synthetic face generation, behavioral mimicry, or physical adversarial examples, has become a vital research direction.

The planned workshop will provide a platform to explore how synthetic data can not only support robust and fair AI model development but also facilitate the simulation, generation, and detection of attacks in face and gesture analysis systems. By fostering collaboration among researchers and practitioners, SD-FGA 2026 seeks to bridge the gap between synthetic data generation and adversarial robustness, ultimately advancing secure, ethical, and scalable AI solutions.

Topics of Interest

Topics of interest include, but are not limited to: 

• Cutting-edge generative techniques for producing realistic face and gesture data

• Synthetic data for attack generation and detection (physical and behavioral)

• Innovative approaches for labeling and annotating synthetic datasets

• Preventing data leakage and enhancing data security in synthetic data pipelines

• Synthetic data for biometric spoofing and anti-spoofing research

• Techniques for enriching and augmenting real datasets using synthetic data

• Synthetic data as a tool for bias reduction and promoting fairness in AI models

• Techniques for enriching and augmenting real datasets using synthetic data

• Synthetic data as a tool for bias reduction and promoting fairness in AI models

• Criteria and methodologies for assessing synthetic data quality and robustness

• Privacy-preserving synthetic data generation for sensitive applications

• New applications in augmented reality, animation, simulation, and virtual environments

• Comparative performance benchmarks and cross-domain evaluation of synthetic datasets