TABLE 1 - uploaded by Zahid Akhtar
Content may be subject to copyright.
Source publication
Despite new technologies that make face detection and recognition more sophisticated, long-recognized problems in security, privacy, and accuracy persist. Refining this technology and introducing it into new domains will require solving these problems through focused interdisciplinary efforts among developers, researchers, and policymakers.
Context in source publication
Context 1
... from their classification cate- gory, FRSs differ according to the face recognition methods they use, which fall roughly into four types. 1,2 Table 1 lists some examples along with the year they first appeared in the liter- ature. 2 FIGURE 1. Tasks in a face recognition system (FRS). The FRS has seven main modules that reflect the tasks it undergoes during (a) the enrollment stage of acquiring, detecting, and normalizing an individual's face and then extracting a feature set, which it then stores as a template in the database. ...
Citations
... Facial recognition has been approached through various methods over the years. Some of those major methods are feature-based (also known as local), holistic, and hybrid matching [7]. ...
... 3) Hybrid: This type of matching method employs both feature-based and holistic matching methods on 3D face images [7]. ...
... Methods use statistical proof as a template for detecting eyes using eye patch photos as a template, to create a statistical model of a photometric appearance trait. Eye detection using learning-based methods is a promising solution [7,8]. Learning-based methods require a large amount of annotated training data to cover a large of eye appearances. ...
The recognition of the eye region from images is a challenging task, particularly when dealing with dark or thick sunglasses that cause reflections and interfere with accurate identification. To address this issue, a novel system called AVOA-MRCNN-OLSTM has been proposed. This system combines Optimization-driven Long Short-Term Memory (LSTM) with Mask RCNN to achieve precise eye recognition even in the presence of eyeglass frame interference. A mean histogram equalization approach is used in the system's first stage to eliminate noise, which improves the image quality. The system then uses Mask RCNN for segmentation and localization. A potent deep learning model called Mask RCNN can precisely recognize and isolate particular items inside an image. It is used in this instance to identify and divide the eye region. The AVOA-MRCNN-OLSTM framework makes use of LSTM, a recurrent neural network variety that can retain patterns for longer periods. It can efficiently acquire and use temporal information to increase eye recognition accuracy by integrating LSTM into the system. The proposed AVOA-MRCNN-OLSTM system's effectiveness is shown by experimental findings. It outperforms the performance of existing algorithms, achieving a remarkable accuracy of 99% in just 0.02 seconds of computing time. The potential uses of this development include biometric identity, surveillance systems, and human-computer interfaces, all of which need precise eye recognition.
... To avoid being tracked by CCTV, for example, when captured by law enforcement and subsequently escaping from their custody, the vast majority of criminal suspects choose to cover some of their faces with hats or masks [14]. Several researchers have studied face occlusion technology for quite some time [15] with an eye on meeting the needs of real-world security scenarios [16,17], and have made several different attempts to make the technology more user-friendly, as the ethical use of face recognition in areas such as law enforcement investigations requires a set of clear criteria to ensure that this technology is trustworthy and safe [18]. Deep forgery detection techniques are learning-based systems that rely on data to a certain degree. ...
The paper presents an evaluation of a Pareto-optimized FaceNet model with data preprocessing techniques to improve the accuracy of face recognition in the era of mask-wearing. The COVID-19 pandemic has led to an increase in mask-wearing, which poses a challenge for face recognition systems. The proposed model uses Pareto optimization to balance accuracy and computation time, and data preprocessing techniques to address the issue of masked faces. The evaluation results demonstrate that the model achieves high accuracy on both masked and unmasked faces, outperforming existing models in the literature. The findings of this study have implications for improving the performance of face recognition systems in real-world scenarios where mask-wearing is prevalent. The results of this study show that the Pareto optimization allowed improving the overall accuracy over the 94% achieved by the original FaceNet variant, which also performed similarly to the ArcFace model during testing. Furthermore, a Pareto-optimized model no longer has a limitation of the model size and is much smaller and more efficient version than the original FaceNet and derivatives, helping to reduce its inference time and making it more practical for use in real-life applications.
... On the one hand, the face image itself, as a kind of identity information, can be used to identify individuals, and it is necessary to prevent malicious collection and abuse. On the other hand, various personal information such as age, gender, race, facial disability, health status, emotion, and even kinship can be analyzed from face datasets through image processing and data mining algorithms [14]. ...
... However, numerous studies have conducted public surveys about FRT application, indicating their concerns about privacy invasion. In many cases, their facial information is collected involuntarily [42], which may lead to undesirable results of intrusions of privacy [43]. The privacy concerns are affected by privacy control, which means giving users the autonomy to control their private information [44]. ...
The application of facial recognition technology (FRT) can effectively reduce the red-light running behavior of e-bikers. However, the privacy issues involved in FRT have also attracted widespread attention from society. This research aims to explore the public and the traffic police's attitudes toward FRT to optimize the use and implementation of FRT. A structured questionnaire survey of 270 people and 94 traffic police in Fuzhou, China, was used. In the study, we use several methods to analyze the investigation data, including the Mann-Whitney U test, Kruskal-Wallis test and Multiple Correspondence Analysis. The survey results indicate that the application of FRT has a significant effect on reducing red-light running behavior. The public’s educational level and driving license status are the most influential factors related to their attitudes to FRT (p < 0.001). The public with these attributes shows more supportive attitudes to FRT and more concerns about privacy invasion. Besides, there are significant differences between the public and the traffic police in attitudes toward FRT (p < 0.001). Compared with the public, the traffic police officers had more supportive attitudes to FRT. This research contributes to promoting the application of FRT legitimately and alleviating people's concerns about the technology.
... As stated by many works in the literature [2], illumination changes, face pose, occlusion and different facial expressions are the main challenges which can be faced in real applications and that should be addressed. Many approaches of the literature have achieved good face recognition performance thanks to the extraction of discriminant features, which constitutes one of the key stages in the overall recognition framework. ...
The evaluation of face recognition algorithms relies on the diversity of the challenges simulated on the adopted benchmarks. The main face recognition challenges cover the illumination changes, inhomogeneous background, different facial expressions, pose variations, occlusion, aging and resolution. Many 2D databases that include one challenge or more have been proposed in the state of the art. These databases represent different amount of individuals and samples, and generally the number of persons does not exceed 100 classes. The reason behind this limitation relies on the resources and materials required to construct a database composed of thousands of images and hundreds of persons along with the mentioned face recognition basic challenges. As a solution, researchers proposed to build benchmarks based on collecting the web images of celebrities from search engines such as Google Images and Flicker. The well-known database of this kind is Labeled Faces in the Wild (LFW) as a public benchmark for face verification. This solution managed to constitute a dependent way to construct benchmarks, but it could not be applicable for face recognition since the collected images have a low resolution and the majority of the persons are represented over few samples (one or two in most cases), which made these databases extremely hard for handcrafted-based face recognition systems. In this paper, we propose to construct a challenging database referred to as mixed face recognition database (MFRD) based on gathering the images of eight well-known benchmarks of the literature (FERET, Extended Yale B, ORL, AR, FEI, KDEF, IMM and JAFFE). The constructed database is expected to be more complex in terms of the amount of classes/images and the diversity of challenges. We expect then that the recognition performance on this database will drop compared to the one recorded on each considered benchmark individually. This paper presents also a new LBP variant, namely dual neighborhood thresholding patterns based on directional sampling (DNTPDS) as a robust and computationally efficient handcrafted descriptor for face recognition. The concept behind this new descriptor is based on defining a 5×5\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$5\times 5$$\end{document} neighborhood topology, that relies on a directional sampling to select the only 16 prominent neighbors instead of 25. The proposed DNTPDS operator demonstrates a superior performance and outperforms 18 state-of-the-art LBP variants that is proved through a set of comprehensive experiments.
... Computational intelligence in face recognition based on visible images has been rapidly developed and widely used in many practical scenarios, such as security [1][2][3][4][5][6][7][8][9], finance [10][11][12][13][14][15][16][17][18][19], and health care service [20][21][22][23][24][25][26][27][28][29]. However, limited by lighting conditions [30][31][32][33], facial expression [34,35], pose [36,37], occlusion [38,39], and other confounding factors [40], some valuable information are easily to be lost. ...
Face recognition has been rapidly developed and widely used. However, there is still considerable uncertainty in the computational intelligence based on human-centric visual understanding. Emerging challenges for face recognition are resulted from information loss. This study aims to tackle these challenges with a broad learning system (BLS). We integrated two models, IR³C with BLS and IR³C with a triplet loss, to control the learning process. In our experiments, we used different strategies to generate more challenging datasets and analyzed the competitiveness, sensitivity, and practicability of the proposed two models. In the model of IR³C with BLS, the recognition rates for the four challenging strategies are all 100%. In the model of IR³C with a triplet loss, the recognition rates are 94.61%, 94.61%, 96.95%, 96.23%, respectively. The experiment results indicate that the proposed two models can achieve a good performance in tackling the considered information loss challenges from face recognition.
... As a future work, we aspire to find the best set of parameters that can guarantee the quality of the findings on larger datasets. Also, we will study the robustness of the frameworks against different adversaries [1,2]. ...
These days, automated face recognition systems are hugely being applied in diverse applications ranging from personal use to border crossing. Feature extraction/representation is extremely vital module in any biometric systems, including face recognition. Thus, the main contribution of this paper is the proposition of a novel descriptor based on monogenic signal representation and Binarized Statistical Image Feature (BSIF) to extract quite distinctive relevant features from face image, named (M-BSIF). In fact, BSIF has not always efficient for face feature extraction, as it was not able to attain the best recognition rates. In order to enhance the capability of BSIF feature representation, our proposed feature description scheme, first applies band pass mechanism via log-Gabor filter on the image, then a monogenic filter is applied to decompose face image into three complementary parts, i.e., local amplitude, local phase, and local orientation. Next, BSIF is utilized to encode these complementary components in order to extract M-BSIF features. Experimental analyses on three publicly available databases (i.e., ORL database, AR database and JAFFE database) demonstrate the efficacy of the proposed M-BSIF descriptor. The proposed system outperforms a framework using only single BSIF.
... As twins have very similar looks, expressions, features, textures, etc. Even sometimes humans also fail to recognize the faces of twins [81]. ...
Unleashing the Art of Digital Forensics is intended to describe and explain the steps taken during a forensic examination, with the intent of making the reader aware of the constraints and considerations that apply during a forensic examination in law enforcement and in the private sector.
Key Features:
• Discusses the recent advancements in Digital Forensics and Cybersecurity
• Reviews detailed applications of Digital Forensics for real-life problems
• Addresses the challenges related to implementation of Digital Forensics and Anti-Forensic approaches
• Includes case studies that will be helpful for researchers
• Offers both quantitative and qualitative research articles, conceptual papers, review papers, etc.
• Identifies the future scope of research in the field of Digital Forensics and Cybersecurity.
This book is aimed primarily at and will be beneficial to graduates, postgraduates, and researchers in Digital Forensics and Cybersecurity.
... Privacy and Security: The widespread use of FRT enables the creation of detailed databases about people's actions and whereabouts, raising a host of concerns about control over personal information and the uses to which it is put, especially when the technology is used by authoritarian governments and/or for commercial interests. Further, demographic and other private information can be revealed by unique facial features, leading to functional creep and intrusions of privacy [15]. As FRT develops, it is likely to treat people's faces not just as a form of biometric identification, but also as a new source of demographic and psychographic data. ...
Facial recognition technology (FRT) is now being introduced across various aspects of public life. However, the controversial nature of FRT and improper uses often generate critical concerns and even resistance. Research on human interactions with FRT has focused principally on individual-level usage in private spaces, tending not to capture in-situ, nuanced human-surveillance technology interactions. To address this gap, we investigated users’ lived experiences with a facial recognition system at a university in the United States, using semi-structured interviews. In this paper, we reported findings of participants’ first impressions and initial reactions to FRT, whether and why their attitudes changed afterwards, and how they evaluated the administration that made the deployment decision. We found that besides issues of privacy, data security, and possible bias, the participants highlighted the idea that FRT might deconstruct the nature of community and connections between people as well as resulting in mass surveillance. In evaluating the deployment decision, the participants perceived control of and transparency in the decision-making process, the accuracy and timeliness of the information, and respect accorded to users in the process as equal in importance to the technology itself. Our findings also point to organizational issues associated with the administration of FRT and offer insights into controversial technology deployment from an organizational justice perspective.
