Synthetic Face Images Research Articles

Can You Tell Real from Fake Face Images? Perception of Computer-Generated Faces by Humans

With recent advances in machine learning and big data, it is now possible to create synthetic images that look real. Face generation is often of particular interest, as faces can be used for various purposes. However, improper use of such content can lead to the dissemination of false information, such as fake news, and thus pose a threat to society. This work studies whether people believe the truthfulness of faces using eye tracking and self-reports, including free-form textual explanations when participants encounter real and computer-generated faces. We used three different datasets for our evaluations, and our experimental results show that while people are relatively better at identifying the truthfulness of real faces and faces generated by earlier machine learning algorithms with different gazing behaviors in viewing and rating phases, they perform less accurately when deciding the truthfulness of synthetic face images that are generated by newer algorithms. Our findings provide important insights for society and policymakers.

Ethics-aware face recognition aided by synthetic face images

Current face recognition models trained on large-scale face datasets have achieved promising performance. However, using face images to train a face recognition model without consent would lead to severe privacy and ethical issues. Moreover, existing face recognition models also exhibit uneven performance on different races, thus perplexing vulnerable populations. To address the aforementioned two issues, this work investigates an ethics-aware face recognition method and examines whether we can leverage synthesized faces to achieve a high-accuracy racial balanced recognition model. In a nutshell, we introduce a race-controllable and identity-innumerable face synthesis approach to generate synthetic face images, and then employ the synthesized images to improve face recognition accuracy and mitigate recognition imbalance among different races despite the scarcity of consenting images (less than 100 individuals). More importantly, the synthetic data enable us to analyze the potential impacts of races on face recognition models quantitatively and facilitate the eradication of racial imbalance in face recognition. Extensive experiments demonstrate that employing our synthetic face data improves face recognition accuracy by a large margin while mitigating the recognition imbalance across different race groups.

Data-agnostic Face Image Synthesis Detection using Bayesian CNNs

Face image synthesis detection is considerably gaining attention because of the potential negative impact on society that this type of synthetic data brings. In this paper, we propose a data-agnostic solution to detect the face image synthesis process. Specifically, our solution is based on an anomaly detection framework that requires only real data to learn the inference process. It is therefore data-agnostic in the sense that it requires no synthetic face images. The solution uses the posterior probability with respect to the reference data to determine if new samples are synthetic or not. Our evaluation results using different synthesizers show that our solution is very competitive against the state-of-the-art, which requires synthetic data for training.

A siamese-based verification system for open-set architecture attribution of synthetic images

Despite the wide variety of methods developed for synthetic image attribution, most of them can only attribute images generated by models or architectures included in the training set and do not work with unknown architectures, hindering their applicability in real-world scenarios. In this paper, we propose a verification framework that relies on a Siamese Network to address the problem of open-set attribution of synthetic images to the architecture that generated them. We consider two different settings. In the first setting, the system determines whether two images have been produced by the same generative architecture or not. In the second setting, the system verifies a claim about the architecture used to generate a synthetic image, utilizing one or multiple reference images generated by the claimed architecture. The main strength of the proposed system is its ability to operate in both closed and open-set scenarios so that the input images, either the query and reference images, can belong to the architectures considered during training or not. Experimental evaluations encompassing various generative architectures such as GANs, diffusion models, and transformers, focusing on synthetic face image generation, confirm the excellent performance of our method in both closed and open-set settings, as well as its strong generalization capabilities.

SPLiT: Single Portrait Lighting Estimation via a Tetrad of Face Intrinsics.

This paper proposes a novel pipeline to estimate a non-parametric environment map with high dynamic range from a single human face image. Lighting-independent and -dependent intrinsic images of the face are first estimated separately in a cascaded network. The influence of face geometry on the two lighting-dependent intrinsics, diffuse shading and specular reflection, are further eliminated by distributing the intrinsics pixel-wise onto spherical representations using the surface normal as indices. This results in two representations simulating images of a diffuse sphere and a glossy sphere under the input scene lighting. Taking into account the distinctive nature of light sources and ambient terms, we further introduce a two-stage lighting estimator to predict both accurate and realistic lighting from these two representations. Our model is trained supervisedly on a large-scale and high-quality synthetic face image dataset. We demonstrate that our method allows accurate and detailed lighting estimation and intrinsic decomposition, outperforming state-of-the-art methods both qualitatively and quantitatively on real face images.

SPRITZ-PS: validation of synthetic face images using a large dataset of printed documents

SPRITZ-PS: validation of synthetic face images using a large dataset of printed documents

Comprehensive Vulnerability Evaluation of Face Recognition Systems to Template Inversion Attacks Via 3DFace Reconstruction.

In this paper, we comprehensively evaluate the vulnerability of state-of-the-art face recognition systems to template inversion attacks using 3D face reconstruction. We propose a new method (called GaFaR) to reconstruct 3D faces from facial templates using a pretrained geometry-aware face generation network, and train a mapping from facial templates to the intermediate latent space of the face generator network. We train our mapping with a semi-supervised approach using real and synthetic face images. For real face images, we use a generative adversarial network (GAN)-based framework to learn the distribution of generator intermediate latent space. For synthetic face images, we directly learn the mapping from facial templates to the generator intermediate latent code. Furthermore, to improve the success attack rate, we use two optimization methods on the camera parameters of the GNeRF model. We propose our method in the whitebox and blackbox attacks against face recognition systems and compare the transferability of our attack with state-of-the-art methods across other face recognition systems on the MOBIO and LFW datasets. We also perform practical presentation attacks on face recognition systems using the digital screen replay and printed photographs, and evaluate the vulnerability of face recognition systems to different template inversion attacks. The project page is available at https://www.idiap.ch/paper/gafar.

The Value of AI Guidance in Human Examination of Synthetically-Generated Faces

Face image synthesis has progressed beyond the point at which humans can effectively distinguish authentic faces from synthetically-generated ones. Recently developed synthetic face image detectors boast ``better-than-human'' discriminative ability, especially those guided by human perceptual intelligence during the model's training process. In this paper, we investigate whether these human-guided synthetic face detectors can assist non-expert human operators in the task of synthetic image detection when compared to models trained without human-guidance. We conducted a large-scale experiment with more than 1,560 subjects classifying whether an image shows an authentic or synthetically-generated face, and annotating regions supporting their decisions. In total, 56,015 annotations across 3,780 unique face images were collected. All subjects first examined samples without any AI support, followed by samples given (a) the AI's decision (``synthetic'' or ``authentic''), (b) class activation maps illustrating where the model deems salient for its decision, or (c) both the AI's decision and AI's saliency map. Synthetic faces were generated with six modern Generative Adversarial Networks. Interesting observations from this experiment include: (1) models trained with human-guidance, which are also more accurate in our experiments, offer better support to human examination of face images when compared to models trained traditionally using cross-entropy loss, (2) binary decisions presented to humans results in their better performance than when saliency maps are presented, (3) understanding the AI's accuracy helps humans to increase trust in a given model and thus increase their overall accuracy. This work demonstrates that although humans supported by machines achieve better-than-random accuracy of synthetic face detection, the approaches of supplying humans with AI support and of building trust are key factors determining high effectiveness of the human-AI tandem.

A Hierarchical Approach toward Prediction of Human Biological Age from Masked Facial Image Leveraging Deep Learning Techniques

The lifestyle of humans has changed noticeably since the contagious COVID-19 disease struck globally. People should wear a face mask as a protective measure to curb the spread of the contagious disease. Consequently, real-world applications (i.e., electronic customer relationship management) dealing with human ages extracted from face images must migrate to a robust system proficient to estimate the age of a person wearing a face mask. In this paper, we proposed a hierarchical age estimation model from masked facial images in a group-to-specific manner rather than a single regression model because age progression across different age groups is quite dissimilar. Our intention was to squeeze the feature space among limited age classes so that the model could fairly discern age. We generated a synthetic masked face image dataset over the IMDB-WIKI face image dataset to train and validate our proposed model due to the absence of a benchmark masked face image dataset with real age annotations. We somewhat mitigated the data sparsity problem of the large public IMDB-WIKI dataset using off-the-shelf down-sampling and up-sampling techniques as required. The age estimation task was fully modeled like a deep classification problem, and expected ages were formulated from SoftMax probabilities. We performed a classification task by deploying multiple low-memory and higher-accuracy-based convolutional neural networks (CNNs). Our proposed hierarchical framework demonstrated marginal improvement in terms of mean absolute error (MAE) compared to the one-off model approach for masked face real age estimation. Moreover, this research is perhaps the maiden attempt to estimate the real age of a person from his/her masked face image.

More Real Than Real: A Study on Human Visual Perception of Synthetic Faces [Applications Corner

Deep fakes became extremely popular in the last years, also thanks to their increasing realism. Therefore, there is the need to measures human's ability to distinguish between real and synthetic face images when confronted with cutting-edge creation technologies. We describe the design and results of a perceptual experiment we have conducted, where a wide and diverse group of volunteers has been exposed to synthetic face images produced by state-of-the-art Generative Adversarial Networks (namely, PG-GAN, StyleGAN, StyleGAN2). The experiment outcomes reveal how strongly we should call into question our human ability to discriminate real faces from synthetic ones generated through modern AI.

Enhanced IPCGAN-Alexnet model for new face image generating on age target

Cross aging face recognition ability will decrease to recognize someone's face after a certain time. Adding synthetics face images at a certain age generated from face aging architecture is one way to increase the performance of cross aging face recognition. A synthetics face image can create use the Generative Adversarial Network-based architecture. The current Generative Adversarial Network-based in face aging still needs high computation to create a model. Based on that reason we proposed a new optimal variant of Identity Preserving Conditional Generative Adversarial Network (IPCGAN), to generate a synthetic face image at certain age groups. In the proposed architecture, change made at the structure in the generator module, age classification module, and change the objective function to increasing the accuracy performance when generates a realistic synthetic face image in certain age groups and also speed up the training time. Modification in the age classifier at the proposed network forces our architecture to generate better synthetics face in certain age groups. Evaluation using Facenet, and Age prediction shows our method accuracy has 4.2% better results in k-NN classification, 3.6% better accuracy results in SVM classification, 8.6% better accuracy result in age verification, and 4.5% fewer accuracy results in age prediction.

Deep relightable appearance models for animatable faces

We present a method for building high-fidelity animatable 3D face models that can be posed and rendered with novel lighting environments in real-time. Our main insight is that relightable models trained to produce an image lit from a single light direction can generalize to natural illumination conditions but are computationally expensive to render. On the other hand, efficient, high-fidelity face models trained with point-light data do not generalize to novel lighting conditions. We leverage the strengths of each of these two approaches. We first train an expensive but generalizable model on point-light illuminations, and use it to generate a training set of high-quality synthetic face images under natural illumination conditions. We then train an efficient model on this augmented dataset, reducing the generalization ability requirements. As the efficacy of this approach hinges on the quality of the synthetic data we can generate, we present a study of lighting pattern combinations for dynamic captures and evaluate their suitability for learning generalizable relightable models. Towards achieving the best possible quality, we present a novel approach for generating dynamic relightable faces that exceeds state-of-the-art performance. Our method is capable of capturing subtle lighting effects and can even generate compelling near-field relighting despite being trained exclusively with far-field lighting data. Finally, we motivate the utility of our model by animating it with images captured from VR-headset mounted cameras, demonstrating the first system for face-driven interactions in VR that uses a photorealistic relightable face model.

VIPPrint: Validating Synthetic Image Detection and Source Linking Methods on a Large Scale Dataset of Printed Documents.

The possibility of carrying out a meaningful forensic analysis on printed and scanned images plays a major role in many applications. First of all, printed documents are often associated with criminal activities, such as terrorist plans, child pornography, and even fake packages. Additionally, printing and scanning can be used to hide the traces of image manipulation or the synthetic nature of images, since the artifacts commonly found in manipulated and synthetic images are gone after the images are printed and scanned. A problem hindering research in this area is the lack of large scale reference datasets to be used for algorithm development and benchmarking. Motivated by this issue, we present a new dataset composed of a large number of synthetic and natural printed face images. To highlight the difficulties associated with the analysis of the images of the dataset, we carried out an extensive set of experiments comparing several printer attribution methods. We also verified that state-of-the-art methods to distinguish natural and synthetic face images fail when applied to print and scanned images. We envision that the availability of the new dataset and the preliminary experiments we carried out will motivate and facilitate further research in this area.

IE-WGAN: A model with Identity Extracting for Face Frontal Synthesis

Face pose affects the effect of the frontal face synthesis, we propose a model used for frontal face synthesis based on WGAN-GP. This model includes identity extracting module, which is used to supervise the training of the face generation module. On the one hand, the model improves the quality of synthetic face images, on the other hand, it can accelerate the convergence speed of network training. We conduct verification experiments on CelebA data sets, and the results show that this model improves the graphic quality of frontal synthesis.

Conditional adversarial consistent identity autoencoder for cross-age face synthesis

Learning-based face aging/rejuvenation has witnessed rapid progress in recent years. However, existing methods still suffer from the loss of personalized identity information when synthesizing cross-age faces. In this paper, we propose a Conditional Adversarial Consistent Identity AutoEncoder (CACIAE) to revisit this problem. Firstly, a Res-Encoder is designed to better generate powerful face representation. Secondly, the rectangular kernel is introduced into the encoder to make full use of horizontal continuous characteristic information of faces and to make the synthetic face images more natural. Thirdly, a novel consistent identity loss is proposed to learn more face details and produce more natural identity-preserving images. Further, two discriminators are designed to enforce the generator to generate more realistic and more age-accurate images. Experimental results prove the effectiveness of the proposed method, both qualitatively and quantitatively. The code is available at https://github.com/XH-B/CACIAE.

Blind face images deblurring with enhancement

Face images deblurring has achieved advanced development; however, existing methods involve high computational cost problems. Furthermore, the recovered face images by current methods have the problems of over-smooth textures, ringing artifacts, and poor details. We consider the problem of face images deblurring as a semantic generation task. In this paper, we propose a generative adversarial network (GAN), which includes a perception-inspired blurry removal generator and a discriminator. The proposed generator reconstructs the latent deblurred image by a U-net based network that contains an enhancement module. Face images are highly structured, and thus can be served as a class-specific prior. Considering this, we propose a perceptual loss function to regularize the recovery of face images, which introduces more clear details and reduces the effects of artifacts. The proposed method has a robust capability of generating realistic face images with pleasant visual effects. Extensive experiments on both synthetic and real-world face images demonstrate that the proposed method is comparable with state-of-the-art methods.

Accurate and robust neural networks for face morphing attack detection

Artificial neural networks tend to use only what they need for a task. For example, to recognize a rooster, a network might only considers the rooster’s red comb and wattle and ignores the rest of the animal. This makes them vulnerable to attacks on their decision making process and can worsen their generality. Thus, this phenomenon has to be considered during the training of networks, especially in safety and security related applications. In this paper, we propose neural network training schemes, which are based on different alternations of the training data, to increase robustness and generality. Precisely, we limit the amount and position of information available to the neural network for the decision making process and study their effects on the accuracy, generality, and robustness against semantic and black box attacks for the particular example of face morphing attacks. In addition, we exploit layer-wise relevance propagation (LRP) to analyze the differences in the decision making process of the differently trained neural networks. A face morphing attack is an attack on a biometric facial recognition system, where the system is fooled to match two different individuals with the same synthetic face image. Such a synthetic image can be created by aligning and blending images of the two individuals that should be matched with this image. We train neural networks for face morphing attack detection using our proposed training schemes and show that they lead to an improvement of robustness against attacks on neural networks. Using LRP, we show that the improved training forces the networks to develop and use reliable models for all regions of the analyzed image. This redundancy in representation is of crucial importance to security related applications.

Generative Adversarial Ensemble Learning for Face Forensics

The recent advance of synthetic image generation and manipulation methods allows us to generate synthetic face images close to real images. On the other hand, the importance of identifying the synthetic face images increases more and more to protect personal privacy from those. Although some deep learning-based image forensic methods have been developed recently, it is still challenging to distinguish synthetic images generated by recent image generation and manipulation methods such as the deep fake, face2face, and face swap. To resolve this challenge, we propose a novel generative adversarial ensemble learning method. We train multiple discriminative and generative networks based on the adversarial learning. Compared to the conventional adversarial learning, our method is however more focused on improving the discrimination ability rather than image generation one. To this end, we improve the discriminabilty by ensembling outputs from different two discriminators. In addition, we train two generators in order to generate general and hard synthetic images. By ensemble learning of all the generators and discriminators, we improve the discriminators by using the generated synthetic face images, and improve the generators by passing the combined feedback of the discriminators. On the FaceForensics benchmark challenge, we thoroughly evaluate our methods by comparing the recent methods. We also provide the ablation study to prove the effectiveness and usefulness of our method.

A DCNN-Based Fast NIR Face Recognition System Robust to Reflected Light From Eyeglasses

Due to an increasing need for face recognition under poor lighting conditions, near infrared (NIR) face recognition based on deep convolutional neural networks (DCNN) has become an active area of research. However, in NIR face images of eyeglasses wearers, reflected light is generated around the eyes due to active NIR light sources, and it is one of the main contributors to performance degradation in NIR face recognition. In addition, there have to date been no attempts to lighten DCNN models for NIR face recognition. To solve these problems, we propose a DCNN-based fast NIR face recognition system which is robust to reflected light. This work has two main contributions: 1) We generated synthetic face images of individuals with and without eyeglasses using our proposed CycleGAN-based Glasses2Non-glasses (G2NG) data augmentation. We then constructed an augmented training database by adding the synthetic images, and the database helps to make the NIR face recognition system robust against reflected light. 2) A lightweight NIR FaceNet (LiNFNet) architecture was developed to reduce the computational complexity of the proposed system by adapting the depthwise separable convolutions and linear bottlenecks to VGGNet 16. The proposed architecture reduces the computation required, while improving the performance of NIR face recognition. Through the experiments reported in this paper, we verified that the proposed G2NG data augmentation improved the face recognition validation rate by 99.09% for NIR face images which have the reflected light from eyeglasses. Also, LiNFNet reduces the number of multiplication operations by 4.4×10 9 compared with VGGNet 16.

A paired sparse representation model for robust face recognition from a single sample

Sparse representation-based classification (SRC) has been shown to achieve a high level of accuracy in face recognition (FR). However, matching faces captured in unconstrained video against a gallery with a single reference facial still per individual typically yields low accuracy. For improved robustness to intra-class variations, SRC techniques for FR have recently been extended to incorporate variational information from an external generic set into an auxiliary dictionary. Despite their success in handling linear variations, non-linear variations (e.g., pose and expressions) between probe and reference facial images cannot be accurately reconstructed with a linear combination of images in the gallery and auxiliary dictionaries because they do not share the same type of variations. In order to account for non-linear variations due to pose, a paired sparse representation model is introduced allowing for joint use of variational information and synthetic face images. The proposed model, called synthetic plus variational model, reconstructs a probe image by jointly using (1) a variational dictionary and (2) a gallery dictionary augmented with a set of synthetic images generated over a wide diversity of pose angles. The augmented gallery dictionary is then encouraged to pair the same sparsity pattern with the variational dictionary for similar pose angles by solving a newly formulated simultaneous sparsity-based optimization problem. Experimental results obtained on Chokepoint and COX-S2V datasets, using different face representations, indicate that the proposed approach can outperform state-of-the-art SRC-based methods for still-to-video FR with a single sample per person.