# SREFI: Synthesis of Realistic Example Face Images

arXiv: https://arxiv.org/abs/1704.06693 · HTML: https://mlsanity.com/paper/1704.06693/

## Key facts

- Cited by 12 papers in the map (data through 14 Oct 2026).
- First posted on arXiv 21 Apr 2017.

## Built on by (most cited first, 12 of 12)

- [Protecting Celebrities from DeepFake with Identity Consistency Transformer](https://mlsanity.com/paper/2203.01318/) (2022)
- [CYBORG: Blending Human Saliency Into the Loss Improves Deep Learning](https://mlsanity.com/paper/2112.00686/) (2021)
- [Testing Human Ability To Detect Deepfake Images of Human Faces](https://mlsanity.com/paper/2212.05056/) (2022)
- [A Study of the Human Perception of Synthetic Faces](https://mlsanity.com/paper/2111.04230/) (2021)
- [Teaching AI to Teach: Leveraging Limited Human Salience Data Into Unlimited Saliency-Based Training](https://mlsanity.com/paper/2306.05527/) (2023)
- [Reversing the Irreversible: A Survey on Inverse Biometrics](https://mlsanity.com/paper/2401.02861/) (2024)
- [Grains of Saliency: Optimizing Saliency-based Training of Biometric Attack Detection Models](https://mlsanity.com/paper/2405.00650/) (2024)
- [Analyzing the Impact of Shape & Context on the Face Recognition Performance of Deep Networks](https://mlsanity.com/paper/2208.02991/) (2022)
- [Improving Model's Focus Improves Performance of Deep Learning-Based Synthetic Face Detectors](https://mlsanity.com/paper/2303.00818/) (2023)
- [Training Better Deep Learning Models Using Human Saliency](https://mlsanity.com/paper/2410.16190/) (2024)
- [Divisive Decisions: Improving Salience-Based Training for Generalization in Binary Classification Tasks](https://mlsanity.com/paper/2507.17000/) (2025)
- [What's Old is New Again: Classical Dimensionality Reduction for Efficient Saliency-Guided Biometric Attack Detection](https://mlsanity.com/paper/2606.13528/) (2026)

## Abstract (arXiv)

In this paper, we propose a novel face synthesis approach that can generate an arbitrarily large number of synthetic images of both real and synthetic identities. Thus a face image dataset can be expanded in terms of the number of identities represented and the number of images per identity using this approach, without the identity-labeling and privacy complications that come from downloading images from the web. To measure the visual fidelity and uniqueness of the synthetic face images and identities, we conducted face matching experiments with both human participants and a CNN pre-trained on a dataset of 2.6M real face images. To evaluate the stability of these synthetic faces, we trained a CNN model with an augmented dataset containing close to 200,000 synthetic faces. We used a snapshot of this trained CNN to recognize extremely challenging frontal (real) face images. Experiments showed training with the augmented faces boosted the face recognition performance of the CNN.