Search for: All records

Automated face recognition is a widely adopted machine learning technology for contactless identification of people in various processes such as automated border control, secure login to electronic devices, community surveillance, tracking school attendance, workplace clock in and clock out. Using face masks have become crucial in our daily life with the recent world-wide COVID-19 pandemic. The use of face masks causes the performance of conventional face recognition technologies to degrade considerably. The effect of mask-wearing in face recognition is yet an understudied issue. In this paper, we address this issue by evaluating the performance of a number of face recognition models which are tested by identifying masked and unmasked face images. We use six conventional machine learning algorithms, which are SVC, KNN, LDA, DT, LR and NB, to find out the ones which perform best, besides the ones which poorly perform, in the presence of masked face images. Local Binary Pattern (LBP) is utilized as the feature extraction operator. We generated and used synthesized masked face images. We prepared unmasked, masked, and half-masked training datasets and evaluated the face recognition performance against both masked and unmasked images to present a broad view of this crucial problem. We believe that our study is unique in elaborating the mask-aware facial recognition with almost all possible scenarios including half_masked-to-masked and half_masked-to-unmasked besides evaluating a larger number of conventional machine learning algorithms compared the other studies in the literature. 

The prevalent commercial deployment of automated facial analysis systems such as face recognition as a robust authentication method has increasingly fueled scientific attention. Current machine learning algorithms allow for a relatively reliable detection, recognition, and categorization of face images comprised of age, race, and gender. Algorithms with such biased data are bound to produce skewed results. It leads to a significant decrease in the performance of state-of-the-art models when applied to images of gender or ethnicity groups. In this paper, we study the gender bias in facial recognition with gender balanced and imbalanced training sets using five traditional machine learning algorithms. We aim to report the machine learning classifiers which are inclined towards gender bias and the ones which mitigate it. Miss rates metric is effective in finding out potential bias in predictions. Our study utilizes miss rates metric along with a standard metric such as accuracy, precision or recall to evaluate possible gender bias effectively. 

The research aims to evaluate the impact of race in facial recognition across two types of algorithms. We give a general insight into facial recognition and discuss four problems related to facial recognition. We review our system design, development, and architectures and give an in-depth evaluation plan for each type of algorithm, dataset, and a look into the software and its architecture. We thoroughly explain the results and findings of our experimentation and provide analysis for the machine learning algorithms and deep learning algorithms. Concluding the investigation, we compare the results of two kinds of algorithms and compare their accuracy, metrics, miss rates, and performances to observe which algorithms mitigate racial bias the most. We evaluate racial bias across five machine learning algorithms and three deep learning algorithms using racially imbalanced and balanced datasets. We evaluate and compare the accuracy and miss rates between all tested algorithms and report that SVC is the superior machine learning algorithm and VGG16 is the best deep learning algorithm based on our experimental study. Our findings conclude the algorithm that mitigates the bias the most is VGG16, and all our deep learning algorithms outperformed their machine learning counterparts. 

In biometric systems, the process of identifying or verifying people using facial data must be highly accurate to ensure a high level of security and credibility. Many researchers investigated the fairness of face recognition systems and reported demographic bias. However, there was not much study on face presentation attack detection technology (PAD) in terms of bias. This research sheds light on bias in face spoofing detection by implementing two phases. First, two CNN (convolutional neural network)-based presentation attack detection models, ResNet50 and VGG16 were used to evaluate the fairness of detecting imposer attacks on the basis of gender. In addition, different sizes of Spoof in the Wild (SiW) testing and training data were used in the first phase to study the effect of gender distribution on the models’ performance. Second, the debiasing variational autoencoder (DB-VAE) (Amini, A., et al., Uncovering and Mitigating Algorithmic Bias through Learned Latent Structure) was applied in combination with VGG16 to assess its ability to mitigate bias in presentation attack detection. Our experiments exposed minor gender bias in CNN-based presentation attack detection methods. In addition, it was proven that imbalance in training and testing data does not necessarily lead to gender bias in the model’s performance. Results proved that the DB-VAE approach (Amini, A., et al., Uncovering and Mitigating Algorithmic Bias through Learned Latent Structure) succeeded in mitigating bias in detecting spoof faces. 

In the realm of computer security, the username/password standard is becoming increasingly antiquated. Usage of the same username and password across various accounts can leave a user open to potential vulnerabilities. Authentication methods of the future need to maintain the ability to provide secure access without a reduction in speed. Facial recognition technologies are quickly becoming integral parts of user security, allowing for a secondary level of user authentication. Augmenting traditional username and password security with facial biometrics has already seen impressive results; however, studying these techniques is necessary to determine how effective these methods are within various parameters. A Convolutional Neural Network (CNN) is a powerful classification approach which is often used for image identification and verification. Quite recently, CNNs have shown great promise in the area of facial image recognition. The comparative study proposed in this paper offers an in-depth analysis of several state-of-the-art deep learning based-facial recognition technologies, to determine via accuracy and other metrics which of those are most effective. In our study, VGG-16 and VGG-19 showed the highest levels of image recognition accuracy, as well as F1-Score. The most favorable configurations of CNN should be documented as an effective way to potentially augment the current username/password standard by increasing the current method’s security with additional facial biometrics. 

With the growing popularity of smartphones, continuous and implicit authentication of such devices via behavioral biometrics such as touch dynamics becomes an attractive option. Specially, when the physical biometrics are challenging to utilize, and their frequent and continuous usage annoys the user. This paper presents a touchstroke authentication model based on several classification algorithms and compare their performances in authenticating legitimate smartphone users. The evaluation results suggest that it is possible to achieve comparable authentication accuracies with an average accuracy of 91% considering the best performing model. This research is supervised by Dr. Debzani Deb (debd@wssu.edu), Department of Computer Science at Winston-Salem State University, NC. 

In the ever-changing world of computer security and user authentication, the username/password standard is becoming increasingly outdated. Using the same username and password across multiple accounts and websites leaves a user open to vulnerabilities, and the need to remember multiple usernames and passwords feels very unnecessary in the current digital age. Authentication methods of the future need to be reliable and fast, while maintaining the ability to provide secure access. Augmenting traditional username-password standard with face biometric is proposed in the literature to enhance the user authentication. However, this technique still needs an extensive evaluation study to show how reliable and effective it will be under different settings. Local Binary Pattern (LBP) is a discrete yet powerful texture classification scheme, which works particularly well with image classification for facial recognition. The system proposed here strives to examine and test various LBP configurations to determine their image classification accuracy. The most favorable configurations of LBP should be examined as a potential way to augment the current username and password standard by increasing their security with facial biometrics. 

With the growing popularity of smartphones, continuous and implicit authentication of such devices via behavioral biometrics such as touch dynamics becomes an attractive option, especially when the physical biometrics are challenging to utilize, or their frequent and continuous usage annoys the user. However, touch dynamics is vulnerable to potential security attacks such as shoulder surfing, camera attack, and smudge attack. As a result, it is challenging to rule out genuine imposters while only relying on models that learn from real touchstrokes. In this paper, a touchstroke authentication model based on Auxiliary Classifier Generative Adversarial Network (AC-GAN) is presented. Given a small subset of a legitimate user's touchstrokes data during training, the presented AC-GAN model learns to generate a vast amount of synthetic touchstrokes that closely approximate the real touchstrokes, simulating imposter behavior, and then uses both generated and real touchstrokes in discriminating real user from the imposters. The presented network is trained on the Touchanalytics dataset and the discriminability is evaluated with popular performance metrics and loss functions. The evaluation results suggest that it is possible to achieve comparable authentication accuracies with Equal Error Rate ranging from 2% to 11% even when the generative model is challenged with a vast number of synthetic data that effectively simulates an imposter behavior. The use of AC-GAN also diversifies generated samples and stabilizes training. 

We are creating a streamlined way to adapt WebIDs [1], and biometrics [2] to the cyber world. This involves building a user authentication system that enables quick, fast and secure access. It is understood that compared to traditional username and password user authentication, WebIDs are designed to provide such services. Nevertheless, if an intruder either has direct access to the user's computer or somehow gets the unique certificate of the user, important information can be stolen with solely the use of WebIDs. Since biometric data (e.g. fingerprints, iris scanning, etc.) is unique and not easily duplicated, this possibility can be avoided by including biometrics in the authentication process. We also include an enrollment protocol that checks whether a user has a WebID while trying to access a server. If they do, we allow the user access to the server, and if they do not, by accessing their own server, we register the user for a WebID with their permission. Implementing these features in the WebID protocol will greatly enhance user authentication safety.