Personal Image Research Articles

A novel blind tamper detection and localization scheme for multiple faces in digital images

AbstractFace image manipulation detection (FIMD) is a research area of great interest, widely applicable in fields requiring data security and authentication. Existing FIMD schemes aim to identify manipulations in digital face images, but they possess individual strengths and limitations. Most schemes can only detect specific manipulations under certain conditions, leading to variable success rates across different images. The literature lacks emphasis on detecting manipulations involving multiple faces. This paper introduces a novel blind tamper detection and localization scheme specifically designed for multiple faces in digital images. The proposed multiple faces manipulation detection (MFMD) scheme consists of two stages: face detection and selection, and image watermarking. Through extensive experiments, the MFMD scheme's performance has been evaluated on various multiple‐face images, considering embedding capacity, payload, watermarked image quality, time complexity, and manipulation detection ability. The results demonstrate the MFMD scheme's efficacy in detecting different types of manipulations for multiple faces in images. Furthermore, the watermarked images exhibit high visual quality, even when multiple faces are present. The scheme's efficiency recommends it for practical applications, especially in sharing personal images over unsecured networks. This research advances FIMD techniques by addressing the neglected area of multiple‐face manipulation detection. With improved accuracy, faster processing times, and resilience against various manipulations, the MFMD scheme offers valuable capabilities for enhancing data security and authentication in real‐world scenarios.

Enhancing eID card mobile-based authentication through 3D facial reconstruction

In recent years the number of credentials and digital identities needed to access online services is exploding, increasing the risk of fraud and unauthorized access to personal data.Biometric verification techniques can help to build more secure authentication systems; in particular face recognition algorithms have reached near perfect accuracy under ideal conditions. In Europe 60% of population owns an electronic identification (eID) document which contains biometric data, and almost all member states provide tools to verify the identity of a user who holds an eID card; however, the adoption of biometric authentication schemes is still limited. Reasons for this low adoption are privacy concerns, about the storage or transmission of personal images, and the degradation of performance of facial recognition algorithms with 2D images acquired "in the wild" (real world condition with inconsistency in lighting and positioning, unclear or occluded facial features, noise from the background). In this paper a new authentication scheme is proposed, exploiting images acquired "in the wild" and 3D facial reconstruction techniques which can help in improving the quality of the probed images, as changes in lighting or position have limited impact on facial shape. Several experiments have been conducted in order to investigate the performance of the proposed method, using the Bosphorus and the NIST Face Recognition Grand Challenge (FRGC) datasets. In particular, one of the most accurate ML face recognition algorithms has been selected to compare the performance of 2D snapshot data, obtained from 3D back-projection, against 2D uncontrolled images acquired "in the wild". We show that using 3D facial reconstructions significantly improve face recognition performance, allowing for a more secure and robust authentication system.

Selfitis Behavior in Relation to Narcissism, Personality, Self Esteem and Body Image among Male Youth

Selfitis Behavior in Relation to Narcissism, Personality, Self Esteem and Body Image among Male Youth

Precise region semantics‐assisted GAN for pose‐guided person image generation

AbstractGenerating a realistic person's image from one source pose conditioned on another different target pose is a promising computer vision task. The previous mainstream methods mainly focus on exploring the transformation relationship between the keypoint‐based source pose and the target pose, but rarely investigate the region‐based human semantic information. Some current methods that adopt the parsing map neither consider the precise local pose‐semantic matching issues nor the correspondence between two different poses. In this study, a Region Semantics‐Assisted Generative Adversarial Network (RSA‐GAN) is proposed for the pose‐guided person image generation task. In particular, a regional pose‐guided semantic fusion module is first developed to solve the imprecise match issue between the semantic parsing map from a certain source image and the corresponding keypoints in the source pose. To well align the style of the human in the source image with the target pose, a pose correspondence guided style injection module is designed to learn the correspondence between the source pose and the target pose. In addition, one gated depth‐wise convolutional cross‐attention based style integration module is proposed to distribute the well‐aligned coarse style information together with the precisely matched pose‐guided semantic information towards the target pose. The experimental results indicate that the proposed RSA‐GAN achieves a 23% reduction in LPIPS compared to the method without using the semantic maps and a 6.9% reduction in FID for the method with semantic maps, respectively, and also shows higher realistic qualitative results.

Social representation of persons with disabilities in marketing and advertising practice

The article discusses the issues of including the image of persons with disabilities in marketing and advertising practice. The data reflecting the statistics of disability and consumer behavior in the market of goods and services for the disabled and persons with disabilities are presented. The purpose of the study is to analyze the current practice of marketing and advertising activities using the image of disability and persons with disabilities. It is concluded that both advertising and any other forms of marketing initiatives reflect society, and therefore most advertisers at some point create communication products that somehow reflect their target audience. Good examples of images of people with disabilities and a well-thought-out advertising campaign are designed to increase brand loyalty and at the same time make the product more popular with a wide audience.

Camera Contrast Learning for Unsupervised Person Re-Identification

Unsupervised person re-identification (Re-ID) aims at finding the most informative features from unlabeled person datasets. Some recent approaches adopted camera-aware strategies for model training and have thereby achieved highly promising results. However, these methods simultaneously address intra-ID discrepancies of all cameras and require independent learning under each camera, which increases the complexity of algorithm. To resolve this issue, we present a camera contrast learning framework for unsupervised person Re-ID. Our method first proposes a time-based camera contrastive learning module to facilitate model learning. At each iteration, we follow the time contrast principle to select one camera centroid as proxy of each cluster. By enforcing the samples to converge to positive proxies, the correlation between features and cameras can gradually be reduced. Moreover, we design a 3-dimensional attention module to further reduce intra-ID discrepancies caused by background shifts. By re-weighting each feature map element in a spatial-channel order, our module can exactly find identity-invariant semantic cues from regions of interest in person images, no matter how the background change. Experimental results on several popular datasets prove that our work surpasses existing unsupervised person Re-ID approaches to a remarkable extent. The source codes can be found in https://github.com/HongweiZhang97/CCL.

Person Re-Identification Method Based on Dual Descriptor Feature Enhancement.

Person re-identification is a technology used to identify individuals across different cameras. Existing methods involve extracting features from an input image and using a single feature for matching. However, these features often provide a biased description of the person. To address this limitation, this paper introduces a new method called the Dual Descriptor Feature Enhancement (DDFE) network, which aims to emulate the multi-perspective observation abilities of humans. The DDFE network uses two independent sub-networks to extract descriptors from the same person image. These descriptors are subsequently combined to create a comprehensive multi-view representation, resulting in a significant improvement in recognition performance. To further enhance the discriminative capability of the DDFE network, a carefully designed training strategy is employed. Firstly, the CurricularFace loss is introduced to enhance the recognition accuracy of each sub-network. Secondly, the DropPath operation is incorporated to introduce randomness during sub-network training, promoting difference between the descriptors. Additionally, an Integration Training Module (ITM) is devised to enhance the discriminability of the integrated features. Extensive experiments are conducted on the Market1501 and MSMT17 datasets. On the Market1501 dataset, the DDFE network achieves an mAP of 91.6% and a Rank1 of 96.1%; on the MSMT17 dataset, the network achieves an mAP of 69.9% and a Rank1 of 87.5%. These outcomes outperform most SOTA methods, highlighting the significant advancement and effectiveness of the DDFE network.

Pengaruh Brand Personality, Brand Love dan Brand Image Terhadap Brand Advocacy

This study aims to find out how brand personality, brand love and brand image influence brand advocacy (a case study on wardah cosmetics reviewers in the female daily forum). By using a quantitative method, data from this study were collected from consumers or users of some think skincare products. The sampling technique used a non-probability sampling technique with purposive sampling category, namely 108 respondents. The analytical method used in this study is data/instrument quality testing, data quality testing and hypothesis testing with the help of smartPLS 3.0 software. The results of this study indicate that there are only 2 independent variables namely brand love, brand image which have a positive and significant effect on brand advocacy. The results of testing the structural model evaluation are the R-Square value of 0.843 or 84.3% of the brand advocacy variable which is influenced by brand personality, brand love and brand image variables. While the remaining 16.7% is influenced by other variables not used in this study.

A new microarchitecture hardware proposal for the use of the improved template matching method in face similarity detection

Template matching is a technique used in many different areas in the literature to search for a small template image in a larger source image. The execution time of the technique can be quite high depending on the size of the source image, since the search made in the large image is realized pixel by pixel in the form of frames selected in the template size. This situation creates inconveniences in terms of real-time use of the technique. In addition, the results cannot be interpreted correctly due to the fact that there are many fake matches in the region where there are real matches. In this study, the technique was strengthened with a new and simple proposal called neighborhood pool to prevent fake matches. In order to reduce the long scan time to real-time execution time, the reinforced template matching technique was redesigned as an FPGA-based hardware architecture. This new micro-hardware architecture, which includes the neighborhood pool idea, was synthesized and embedded in the FPGA device by time analysis and tested on the device. For the tests of the study, the face matching scenario was chosen and the face image of a particular person was searched in the larger source image. The improved template matching technique in software-based tests reduced the execution time by about 4.5 times compared to the classical template matching method, and about 300 times in hardware testing, and it was observed that the system execution time became compatible with real-time. When the proposed new micro hardware architecture is compared with the CNN and template matching designs in the literature, it is clearly seen that it has a very low hardware requirement compared to them.

Klasifikasi Citra Mengkudu Berdasarkan Perhitungan Jarak Piksel pada Algoritma K-Nearest Neighbour

Noni fruit is included in exported food commodities in Indonesia. The size of noni fruit, based on human vision, generally has varied shapes with distinctive textures and various patterns, so that the process of filtering fruit based on color and shape can be done in large quantities. In this study, K-Nearest Neighbor (KNN) has been implemented as a classification algorithm because it has advantages in classifying images and is resistant to noise. Noni imagery is a personal image taken from a noni garden in the morning and undergoes a background subtraction process. The imagery quality improvement technique uses the Hue Saturation Value (HSV) color feature and the Gray Level Co-Occurrence Matrix (GLCM) characteristic feature. KNN accuracy without features is lower than using HSV and GLCM features. From the experimental results, the highest accuracy was obtained using HSV-GLCM at K is 1 and d is 1, namely 95%, while the lowest accuracy was 55% using KNN only at K is 5 and d is 8.

An Appraisal of the Implications of Deep Fakes: The Need for Urgent International Legislations

Purpose: Artificial intelligence as a subfield of the Fourth Industrial Revolution has been a controversial concept since its inception. The reason for its controversy is that through it, incredible inventions and also inventions detrimental to the society have surfaced. A deep fake is an Artificial Intelligence technology that creates videos and images of persons and events that in fact did not happen. The aim of this paper is to evaluate the concept of deepfake, its positive impacts and threats posed to individuals, corporate institutions and nations. It further assesses the legal implications of and provisions against deep fakes.
 Methodology: The research methodology adopted in this study is doctrinal. Just like every Artificial Intelligence technology, it has benefits that cut across the marketing, fashion, and art industry among others. However, the issue for determination remains whether the positive impact of this AI technology supersedes the negative impacts.
 Findings: Findings revealed that it has a high tendency to deceive an average person and also create uncertainty about the authenticity of a piece of information. This can lead to the spread of fake news, fraud, blackmail, fabrication of evidence, and even national and global insecurity. Trust in the social media and the internet generally will face a great decline at the spread of deepfakes, this will then be a medium for denial of videos that are in fact true.
 Unique Contribution to Theory, Practice and Policy: The world is set to face a new and modified aspect of social engineering that will require an upgrade in cybersecurity techniques and strategies. By virtue of the sensitivity of the negative impacts of deepfake, this study concludes that it is necessary to promulgate legislations both at national and international levels to curb or regulate deepfakes.

Newton’s Law of Gravitational Force (NLGF) based Machine Learning Technique for Uneven Illuminated Face Detection

A photo gallery is crucial for organizing your photos, presenting them in beautiful categories, and doing sophisticated memory searches. The photo gallery is portrayed in a vocabulary of nonlinear similarities to the prototype face image collection. One of the difficult research ideas for machine learning technologies is the maintenance of a photo gallery using facial recognition. Based on changes in the faces' appearance, faces are identified. This research proposes novel machine learning algorithms to recognize faces by characterizing the majority of discriminating local characteristics, which maximizes the dissimilarity between face photos of different persons and reduces the dissimilarity between features between face images of the same person. This method relies on Newton's third law of gravitational force to determine the relationship between pixels to extract the features of noisy accurately and efficiently, unevenly illuminated, and rotationally invariant face images.

BRPPNet: Balanced privacy protection network for referring personal image privacy protection

Traditional personal image privacy protection usually suffers from the overprotection problem, where one or more undesired persons in an image may be inevitably shielded, yielding unnecessary information loss. Motivated by this, this paper explores a novel task “Referring Personal Image Privacy Protection” (RP-IPP) to protect the designated person in an image according to a user’s text or voice input. We propose a lightweight yet effective personal protection network “Balanced Referring Personal PrivacyNet” (BRPPNet), which introduces a Multi-scale Feature Fusion Module (MFFM) with a proper “Balanced-BCE loss” to effectively localize the referring person. Technically, MFFM adopts a lightweight CNN backbone to filter noise information as well as complement visual features for high-quality mask generation. What is more, we have theoretically proven the insufficiency of binary cross-entropy (BCE) loss and its variants for RP-IPP, which suffers from the serious imbalance problem during gradient propagation, and thus formulate “Balanced-BCE loss” to alleviate the gradient propagation bias caused by unequal positive and negative pixels. To verify the effectiveness of BRPPNet, we manually construct a dataset “Referring Personal COCO” (RPCOCO). The experimental results demonstrate that BRPPNet outperforms the advanced approaches for RP-IPP, and the proposed “Balanced-BCE loss” embedded into several existing approaches consistently boosts performance, yielding remarkable improvements on all the metrics.

Achieving Information Security by multi-Modal Iris-Retina Biometric Approach Using Improved Mask R-CNN

The need for reliable user recognition (identification/authentication) techniques has grown in response to heightened security concerns and accelerated advances in networking, communication, and mobility. Biometrics, defined as the science of recognizing an individual based on his or her physical or behavioral characteristics, is gaining recognition as a method for determining an individual's identity. Various commercial, civilian, and forensic applications now use biometric systems to establish identity. The purpose of this paper is to design an efficient multimodal biometric system based on iris and retinal features to assure accurate human recognition and improve the accuracy of recognition using deep learning techniques. Deep learning models were tested using retinographies and iris images acquired from the MESSIDOR and CASIA-IrisV1 databases for the same person. The Iris region was segmented from the image using the custom Mask R-CNN method, and the unique blood vessels were segmented from retinal images of the same person using principal curvature. Then, in order to aid precise recognition, they optimally extract significant information from the segmented images of the iris and retina. The suggested model attained 98% accuracy, 98.1% recall, and 98.1% precision. It has been discovered that using a custom Mask R-CNN approach on Iris-Retina images improves efficiency and accuracy in person recognition.

Identity Feature Disentanglement for Visible-Infrared Person Re-Identification

Visible-infrared person re-identification (VI-ReID) task aims to retrieve persons from different spectrum cameras (i.e., visible and infrared images). The biggest challenge of VI-ReID is the huge cross-modal discrepancy caused by different imaging mechanisms. Many VI-ReID methods have been proposed by embedding different modal person images into a shared feature space to narrow the cross-modal discrepancy. However, these methods ignore the purification of identity features, which results in identity features containing different modal information and failing to align well. In this article, an identity feature disentanglement method is proposed to disentangle the identity features from identity-irrelevant information, such as pose and modality. Specifically, images of different modalities are first processed to extract shared features that reduce the cross-modal discrepancy preliminarily. Then the extracted feature of each image is disentangled into a latent identity variable and an identity-irrelevant variable. In order to enforce the latent identity variable to contain as much identity information as possible and as little identity-irrelevant information, an ID-discriminative loss and an ID-swapping reconstruction process are additionally designed. Extensive quantitative and qualitative experiments on two popular public VI-ReID datasets, RegDB and SYSU-MM01, demonstrate the efficacy and superiority of the proposed method.

Looks Matter: Are U.S. Schools of Social Work Representing Diversity on Their Websites?

Abstract Public administration researchers have found that passive diversity leads to active diversity. However, whether this relationship holds in schools of social work is uncertain. Understanding this relationship in schools of social work can inform methods used by schools of social work to foster diversity and inclusion. Authors examined the relationship between passive and active diversity representation among 31 randomly selected school of social work websites. Quantitative content analysis revealed that the three most common passive diversity characteristics were (1) having tenure-track faculty of color, (2) images of diverse persons, and (3) faculty with diversity-focused research. The top three active diversity characteristics were (1) no GRE admissions requirement, (2) a woman dean/director, and (3) scholarship/fellowships opportunities. Least represented were student demographics and pronouns in faculty profiles. Most notable bivariate findings were the positive associations between a dean/director of color and tenured faculty of color and the relationship between faculty with diversity-focused research and tenured faculty of color. Greater representation of deans/directors of color in schools of social work is critical for the career advancement of faculty of color. To attract more faculty of color, schools of social work should use their websites to promote hiring opportunities for scholars conducting diversity-focused research. Future research exploring additional diversity characteristics on schools of social work websites and whether these characteristics translate to more inclusive environments is needed.

宋人不言慶曆“新政”“新法”“變法”考———基於觀念史的個案研究

LANGUAGE NOTE | Document text in Chinese; abstract also in English.
 宋人雖常以“新法”“變法”“新政”賅括熙豐政事，但未曾以這些詞彙歸納慶曆之政；當代文史著作常言“慶曆新政”，這説法卻不見於清代以前文獻，原因何在？本文考析宋人慶曆書寫，指出：（1）熙豐變法時期宋人從未直接比較范、王改革主張，也不曾認爲慶曆、熙寧更革是可比的對象，這些聯繫待朱熹而後始。今人或曰熙豐變法之時，慶曆革新派趨於保守，這固屬當代觀點，卻非宋人認識；（2）范仲淹故後，對其顯有貶評者只有梅堯臣、宋神宗、王安石三例，後二者對范的批評，反映熙豐變法主事者並不完全認同慶曆改革；（3）歐陽修等慶曆當事人認爲范仲淹改革失敗皆因“群言營營，卒壞於成”，可是徽宗朝以後卻出現“慶曆之治”説法，其内涵又被簡化爲仁宗能“開納直言，善御群臣，賢必進，邪必退”。這説法無法揭示慶曆之政真正特色，又故意忽略慶曆改革主事者因“衆訾成波，擠落在外”的事實，卻有利於范仲淹作爲忠義第一人形象的建立；（4）宋人從不曾認爲慶曆之政有違祖宗之法，這是宋人不言慶曆“新”政的根本原因。南宋以後論者以范仲淹爲“守我宋之家法者”，進一步確立這觀點。
 When Song-Dynasty people summarized the political situation of the Xining (1068-1077) and Yuanfeng (1078-1085) eras, they used the terms “new laws,” “reform of laws,” and “new policies,” but the Qingli era (1041-1048) was never treated in the same way. The term “New Policies of the Qingli era” is commonly used in contemporary scholarship of Chinese history, but this term was not in existence in the literature before the Qing Dynasty. What is the reason for these phenomena? The present study analyzes Song people’s writing on Qingli and makes the following arguments:
 (1) During the Xining-Yuanfeng reform, Song people had never directly compared the reformist advocacy of Fan Zhongyan (989 1052) and that of Wang Anshi (1021-1086), nor had they ever considered the Qingli and Xining-Yuanfeng reforms comparable, until Zhu Xi (1130-1200) made such a comparison. Today’s view that the Qingli reform advocacy was deemed conservative during the Xining-Yuanfeng reform movement was never so perceived by Song people. (2) After Fan’s death, he was explicitly criticized by Mei Yaochen (1002-1060), Emperor Shenzong of the Song (r. 1067 1085), and Wang Anshi only. The criticisms by the latter two indicate that the Xining-Yuanfeng reform leaders did not completely agree with the Qingli reform. (3) Ouyang Xiu (1007-1072) and the other leaders of the Qingli reform held that Fan’s failure was due to the extensive wave of criticisms the reform had triggered. However, during the reign of Emperor Huizong (r. 1100-1126), a positive view emerged, coming with it inarguable narratives that describe the reform as a sign of Emperor Renzong’s (r. 1022-1063) willingness to accept advice. Unable to reveal the Qingli reform’s true features, this opinion deliberately ignores the historical fact that the Qingli advocates were lambasted and thus demoted, but helps construct Fan’s personal image as a most loyal person. (4) Song people had never considered that the Qingli reform violated the ancestral custom of the Song Dynasty, and this is the very reason why they had never referred to it as “new policies.” This argument was further supported by the opinion of Southern Song critics, who viewed Fan as a protector of the traditional norms laid down by the ancestors of the Song Dynasty’s royal house.

Our Photographs, Old and New

Private and family photographs are our most personal images taken at significant events of our lives. We used to put these photos into albums or just kept them in boxes without specific orders, and we only opened up these personal archives when we wanted to share our memories and stories with family members or close friends. These common habits regarding how we take and share private photos have changed a lot due to the digital turn, the increasing number of digital cameras, and the broadening usage of the Internet and social media platforms. Archives are often places of safeguarding memories, places that make us remember and places of preservation. However, in recent years, the questions related to them and the archiving methods are mostly focused on how digitization has transformed them. The varying ways and effects of technical image usage caused by the technological boom are some of the most characteristic and defining elements of our times. In my paper titled Our Photographs, Old and New, I examine — through a contemporary artwork by Gergely Barcza — how private photography practices and private photo archives have changed in the last decades.

Satisfaction with Service Quality and Re-treatment Intention according to Perception of Eyebrows and Motives for Semi-permanent Makeup Eyebrow Treatment

This study attempted to investigate people’s satisfaction with service quality and re-treatment intention according to the perception of eyebrows, semi-permanent makeup and motives for eyebrow treatment against those in their 20-50s or older, who had semi-permanent makeup or eyebrow treatment before, and the results found the followings: In terms of eyebrow shape, ‘standard’ was the highest with 37.7%, followed by ‘straight (21.7%)’ and ‘angular (18.3%)’. When asked how much satisfied they are with their eyebrow shape, ‘not satisfied (2 points)’ was responded with 2.33 points in average. When asked what bothers them about their eyebrows, ‘loss of eyebrows (18.7%)’ was most responded, followed by ‘unsatisfying shape (18.5%)’. When asked if an eyebrow shape has an influence on their personal image, ‘4.52 points’ was observed in average, being positioned between ‘likely (4 points)’ and ‘very likely (5 points)’. Concerning the effects of the motives to get a semi-permanent makeup or eyebrow treatment on satisfaction with service quality, ‘aesthetic interest’ and ‘pursuit of convenience’ revealed a positive influence with statistical significance. Regarding the effects of satisfaction with such service quality on re-treatment intention, ‘satisfaction with the process’ and ‘satisfaction with the results’ showed a positive influence with statistical significance. In addition, motives to get a semi-permanent makeup or eyebrow treatment, satisfaction with service quality and re-treatment intention were very high with 3.84 points in average. In other words, the above results confirm that motives to get a semi-permanent makeup or eyebrow treatment and satisfaction with service quality have a positive correlation with retreatment intention.

Vision Transformer with hierarchical structure and windows shifting for person re-identification.

Extracting rich feature representations is a key challenge in person re-identification (Re-ID) tasks. However, traditional Convolutional Neural Networks (CNN) based methods could ignore a part of information when processing local regions of person images, which leads to incomplete feature extraction. To this end, this paper proposes a person Re-ID method based on vision Transformer with hierarchical structure and window shifting. When extracting person image features, the hierarchical Transformer model is constructed by introducing the hierarchical construction method commonly used in CNN. Then, considering the importance of local information of person images for complete feature extraction, the self-attention calculation is performed by shifting within the window region. Finally, experiments on three standard datasets demonstrate the effectiveness and superiority of the proposed method.