Facial Analysis Research Articles

A multi-ancestry GWAS meta-analysis of facial features and its application in predicting archaic human features

Facial morphology, a complex trait influenced by genetics, holds great significance in evolutionary research. However, due to limited fossil evidence, the facial characteristics of Neanderthals and Denisovans have remained largely unknown. In this study, we conducted a large-scale multi-ethnic meta-analysis of the genome-wide association study (GWAS), including 9674 East Asians and 10,115 Europeans, quantitatively assessing 78 facial traits using 3D facial images. We identified 71 genomic loci associated with facial features, including 21 novel loci. We developed a facial polygenic score (FPS) that enables the prediction of facial features based on genetic information. Interestingly, the distribution of FPSs among populations from diverse continental groups exhibited relevant correlations with observed facial features. Furthermore, we applied the FPS to predict the facial traits of seven Neanderthals and one Denisovan using ancient DNA and aligned predictions with the fossil records. Our results suggested that Neanderthals and Denisovans likely shared similar facial features, such as a wider but shorter nose and a wider endocanthion distance. The decreased mouth width was characterized specifically in Denisovans. The integration of genomic data and facial trait analysis provides valuable insights into the evolutionary history and adaptive changes in human facial morphology.

Artificial Intelligence-Driven Facial Image Analysis for the Early Detection of Rare Diseases: Legal, Ethical, Forensic, and Cybersecurity Considerations

Artificial Intelligence-Driven Facial Image Analysis for the Early Detection of Rare Diseases: Legal, Ethical, Forensic, and Cybersecurity Considerations

Quality of life assessment in children with dental arch constriction and impaired nasal breathing

Relevance. Malocclusion often cause morphofunctional and aesthetic disturbances in the dentoalveolar system, and the lack of nasal breathing in children exacerbates existing deviations.Purpose. To assess morphofunctional parameters and quality of life indicators in children with impaired nasal breathing.Materials and methods. Photometric facial analysis, cephalometric and anthropometric calculations, anterior active rhinomanometry, and an oral health-related quality of life questionnaire for children (OHRQoL) were conducted.Results. In patients with impaired nasal breathing, social and family well-being parameters were reduced in the quality of life assessment.Conclusion. An orthodontist can determine the causes and degree of nasal breathing impairment by analyzing teleradiographs, orthopantomograms, and studying the quality of life in children. This aids in interdisciplinary planning for early diagnosis and treatment of children with maxillary constriction.

Real-time Driver Drowsiness Detection Techniques using Machine Learning Algorithms and Models

This paper is a report on the research conducted in the field of electronics and communication engineering to develop a system for driver drowsiness detection to prevent accidents happening because of driver fatigue and sleepiness. The document discusses the solution by Tensor flow machine learning tool and open CV image pre-processing available for detecting fatigue and their efficacy in preventing accidents in the current state of traffic. Furthermore, the paper states the overview of the observations made by the authors in order to help further optimization in the mentioned field to achieve the utility at a better efficiency for a safer road. Key Words: Deep Learning, Convolutional Neural Networks (CNNs), Real-time Monitoring, Facial Feature Analysis, Eye Closure Detection, EEG Signal Analysis

Three-dimensional prediction of the nose for facial reconstruction: A preliminary study on North Indian adults

Forensic facial reconstruction is the last recourse to establish the identity of an unknown skull. The facial soft-tissue thickness (FSTT) is required to reconstruct various facial features on a skull. Unlike other facial features, the nose is made of cartilaginous tissue except for a small nasal bone. A large cavity (pyriform aperture) exists on the skull in place of the nose, which makes it a challenging job for reconstruction. The nose is a vital feature for the recognition of a face. Any change in the shape or size of the nose can alter the original aesthetic of the face. The present study proposes angles and regression functions on the bony structure to predict the various parts of the soft nose. A sample of computed tomography (CT) images of 100 males and 100 females aged between 18 and 45 years were included in the study. Apart from measuring fourteen linear parameters with three angles, simple linear regression models were derived for five pairs of parameters. Pearson's correlation coefficients for most of the parameters ranging between 0.221 and 0.872 were found to be significant at p ≤ 0.05 level. FSTT at three anatomical landmarks of the nose was also measured. A morphological observation study was undertaken to find the most frequent direction of the bony anterior nasal spine (ans) and its relation with the position of the pronasale (prn) on the soft nose. The devised parameters proposed in the study may also prove useful for reconstructing the nose in other populations.

Three-Dimensional Anthropometric Facial Analysis and Fitting Discrepancies Between Stereophotogrammetry and CT.

The reliability and repeatability of stereophotogrammetry and CT in 3-dimensional anthropometric facial analysis were investigated in this study, which also explored the clinical application of supine CT. In this study, 3D CT and 3dMD stereophotogrammetry were used on 20 healthy volunteers. The fitting distance between stereophotogrammetry and CT scans at landmark points was measured, along with facial feature measurements (Al-Al) face width (Go-Go, Zy-Zy, Ex-Ex), and hemi-face height (Sn-Gn). The intraclass correlation coefficient (ICC) was employed to assess interrater agreement and to verify the reliability of the measurement methods. Paired t -analysis was utilized for analyzing intramethod displacement. The alare, nasion, and pronasale points were found to be minimally influenced by different positions and are more recommended as landmark points for registration. CT demonstrated good interrater reliability on all indicators. In stereophotogrammetry, measurements for Go-Go and Zy-Zy displayed an interclass correlation coefficient (ICC) of less than 0.75. Significant differences between the 2 methods were observed for En-En, Ex-Ex, and Go-Go ( P < 0.05). Specifically, CT analysis for Go-Go showed a measurement 2.34 mm larger than that obtained with the 3dMD method. Both CT and stereophotogrammetry were found to be reliable methods for evaluating facial soft tissue. It is speculated that Go-Go measurement is primarily influenced by factors such as different positions, facial expressions, and gravity. These variables should be carefully considered during the evaluation of the mandibular angle region.

Decoding faces: Misalignments of gender identification in automated systems

Automated Facial Analysis technologies, predominantly used for facial detection and recognition, have garnered significant attention in recent years. Although these technologies have seen advancements and widespread adoption, biases embedded within systems have raised ethical concerns. This research aims to delve into the disparities of Automatic Gender Recognition systems (AGRs), particularly their oversimplification of gender identities through a binary lens. Such a reductionist perspective is known to marginalize and misgender individuals. This study set out to investigate the alignment of an individual's gender identity and its expression through the face with societal norms, and the perceived difference between misgendering experiences from machines versus humans. Insights were gathered through an online survey, utilizing an AGR system to simulate misgendering experiences. The overarching goal is to shed light on gender identity nuances and guide the creation of more ethically responsible and inclusive facial recognition software.

Facial morphological analysis for the development of a representative atlas of the brazilian population

The objective of this study was to propose categories of morphological classification for the face and its anatomical structures, as well as to propose illustrations to support the development of an atlas that facilitates facial morphological analysis of adult Brazilians. It was a descriptive study based on the analysis of the frequency and distribution of 13 photoanthropometric facial ratios (RFAs) obtained from a representative sample of the Brazilian population. RFAs related to facial height and width, eye width, intercanthal distance, nose length and width, philtrum ridge height and width, mouth thickness and width, upper and lower lip thickness, and chin height were analyzed. The study included a sample of 5.000 individuals aged between 18 and 22 years, evenly distributed between genders. Data normality was assessed using the Shapiro-Wilk test, considering them as parametric when p > 0.05. For the RFAs that showed normal distribution, mean ± 1.5 standard deviations (SD) were used to categorize facial measurements as regular, below average, or above average. Non-parametric RFAs were analyzed based on the median and 10th and 90th percentiles of the data. Based on the established average iris diameter, which is considered the most stable facial measurement, the values of the described RFAs were converted to a numerical scale in centimeters, allowing for the illustration of female and male faces. In this way, it was possible to categorize the facial anatomical structures and, consequently, visualize the facial morphological pattern of the adult Brazilian population.

Being facially expressive is socially advantageous

Individuals vary in how they move their faces in everyday social interactions. In a first large-scale study, we measured variation in dynamic facial behaviour during social interaction and examined dyadic outcomes and impression formation. In Study 1, we recorded semi-structured video calls with 52 participants interacting with a confederate across various everyday contexts. Video clips were rated by 176 independent participants. In Study 2, we examined video calls of 1315 participants engaging in unstructured video-call interactions. Facial expressivity indices were extracted using automated Facial Action Coding Scheme analysis and measures of personality and partner impressions were obtained by self-report. Facial expressivity varied considerably across participants, but little across contexts, social partners or time. In Study 1, more facially expressive participants were more well-liked, agreeable, and successful at negotiating (if also more agreeable). Participants who were more facially competent, readable, and perceived as readable were also more well-liked. In Study 2, we replicated the findings that facial expressivity was associated with agreeableness and liking by their social partner, and additionally found it to be associated with extraversion and neuroticism. Findings suggest that facial behaviour is a stable individual difference that proffers social advantages, pointing towards an affiliative, adaptive function.

Facial Skin Analysis in Malaysians using YOLOv5: A Deep Learning Perspective

Nowadays, people are more concerned about their skin conditions and are more willing to spend money and time on facial care routines. The beauty sector market is increasing, and more skin type readers are being created to help people determine their skin type. While various skin type readers are in the market, each is invented and tested abroad. Those skin type readers in the beauty market are not applied well on Malaysian skin. Therefore, this paper proposes a facial skin analysis system tailored primarily for Malaysian skin. This paper integrated object detection and deep learning algorithms in developing skin-type readers. A unique dataset consisting solely of facial images of Malaysian skin was created from scratch for the model. Additionally, You Only Look Once version 5 (YOLOv5) is employed to detect users' facial skin conditions, such as acne, pigment, enlarged pores, uneven skin, blackheads, etc. Then, based on the detected skin conditions, it further classifies the user's skin type into the normal, oily, sensitive, or dry groups.

Analisis Kesehatan Wajah pada Laki-Laki dan Perempuan Usia Produktif di SMA Kalam Kudus II, Jakarta

Facial analysis plays an important role in early identification of medical conditions through skin changes. Face Analyzer technology offers accurate detection of skin health parameters such as elasticity, hydration, texture and signs of aging. This service activity was carried out at Kalam Kudus High School involving 68 men and women of productive age (18-64 years). Skin measurements using the Skin Analyzer include parameters for pores, spots, wrinkles and roughness. The majority of respondents were aged 18-35 years (52.9%), followed by those aged 36-50 years (27.9%) and 51-64 years (19.1%), with 61.8% female. Data show significant variations in skin parameters based on age, with increased wrinkles and skin roughness at older ages. The results show the importance of early detection and intervention to maintain healthy skin. Regular monitoring and education using advanced technology such as the Face Analyzer can increase awareness and prevention of skin problems. The use of Face Analyzer technology and comprehensive skin health education is very important to improve the skin health of individuals of productive age, prevent skin problems, and improve quality of life.

Donor Limb Functional Restoration via a Novel Clinical Care Pathway following Fibula Free Flap Harvest for Head and Neck Reconstruction.

Patients undergoing head and neck skeletal reconstruction (HNR) often require free tissue transfer from the extremities to ensure proper restoration of form and function. This requires a team-based, highly reliable medical system centered around the patient needs. Surgical intervention across multiple sites and harvesting of donor tissue results in short- and long-term physical impairments. There is a paucity of research objectively measuring impairments resulting from the graft donor site. There is a lack of research that objectively measures impairments and protocols for the management of these patients postoperatively. Patients undergo little, if any, formal approach to dealing with the vast impairments, which are sequelae to this surgery. This leads to large discrepancies in proposed functional progressions, return to duty timelines, and utilization of rehabilitative resources. At a major military medical center, an innovative clinical care pathway for patients undergoing HNR using free tissue transfer was implemented using a multidisciplinary model that focuses on early engagement with rehabilitation. This model, paired with a single surgery, will attempt to return service members to duty months earlier than the traditional approach. This report describes the conceptual framework and implementation of a new criteria-based, multidisciplinary clinical care pathway for HNR patients. The collaboration amongst the multidisciplinary care team has optimized the holistic health of the patient and communication with their support network, yielding faster return to normalization of daily life activities. The long-term goal is to further develop and formalize this pathway to best serve this patient population.

Less is more: Engagement with the content of social media influencers

We draw upon theories of social media engagement to explore the factors affecting the success of the various influencer types, based on the size of their audience. We use the social media content of 8,076 influencers and employ sentiment analysis of text and facial recognition analysis of pictures in their content to examine what drives engagement. We show that the social media content of micro-influencers is more likely to be marked as favourite, while the content of other influencer types is more likely to be shared. We further show that including pictures in the content can result in higher engagement and that showing a person in the pictures also affects engagement, but the strength of this effect depends on the size of the influencer’s audience. Our findings provide novel insights into the theories of social media engagement and sorely needed practical implications regarding content creation on social media platforms.

FACIAL IMAGE CAPTIONING USING DNN

Facial analysis, encompassing emotion, age, and gender detection, shows potential in various applications such as human-computer interaction, business, security, and health. This study delves into the development and evaluation of a deep neural network (DNN) model for facial emotion, age, and gender detection. Utilizing a convolutional neural network (CNN) architecture trained on diverse datasets for each task, our model proves effective in predicting facial features. The accuracy of needs assessment is X%, the marginal error (MAE) of age estimation is Y years, and the accuracy of gender classification is Z%.

Facial Feature-Based Attention Tracking System for Enhanced Online Learning Engagement

Recognizing and enhancing student engagement is crucial for improving learning outcomes, particularly in the context of online classes where monitoring can be challenging. Traditional methods of attendance tracking, such as calling out names, are impractical and susceptible to manipulation in the virtual environment. Students might appear 'online' without actively participating, and the absence of video feeds makes it difficult for teachers to verify attendance and attention. In order to realize a highly efficient and robust attendance management and engagement level prediction system for online learning, In the proposed System, the learner’s face is monitored by a video camera while attending a video lecture. Facial features were analyzed to predict reaction time (RT) to a task-irrelevant stimulus, which was assumed to be an index of the level of attention. Then apply a machine learning method, light Gradient Boosting Machine (LightGBM), to estimate RTs from facial features extracted as action units (AUs) corresponding to facial muscle movements by an open-source software (OpenFace). This project is to develops a user-friendly system integrated with private online learning and attendance recording system for teachers that can automatically record students ‘engagement state and attendance then generate attendance reports for online classrooms. It encompasses a novel design using the AI based FFCNN (Face Fiducial Convolution Neural Network) model to capture face biometric randomly from students’ video stream and record their attendance automatically. This integrated solution not only streamlines attendance management but also provides valuable insights into students' engagement levels through facial feature analysis.

Computer-based facial recognition as an assisting diagnostic tool to identify children with Noonan syndrome

BackgroundNoonan syndrome (NS) is a rare genetic disease, and patients who suffer from it exhibit a facial morphology that is characterized by a high forehead, hypertelorism, ptosis, inner epicanthal folds, down-slanting palpebral fissures, a highly arched palate, a round nasal tip, and posteriorly rotated ears. Facial analysis technology has recently been applied to identify many genetic syndromes (GSs). However, few studies have investigated the identification of NS based on the facial features of the subjects.ObjectivesThis study develops advanced models to enhance the accuracy of diagnosis of NS.MethodsA total of 1,892 people were enrolled in this study, including 233 patients with NS, 863 patients with other GSs, and 796 healthy children. We took one to 10 frontal photos of each subject to build a dataset, and then applied the multi-task convolutional neural network (MTCNN) for data pre-processing to generate standardized outputs with five crucial facial landmarks. The ImageNet dataset was used to pre-train the network so that it could capture generalizable features and minimize data wastage. We subsequently constructed seven models for facial identification based on the VGG16, VGG19, VGG16-BN, VGG19-BN, ResNet50, MobileNet-V2, and squeeze-and-excitation network (SENet) architectures. The identification performance of seven models was evaluated and compared with that of six physicians.ResultsAll models exhibited a high accuracy, precision, and specificity in recognizing NS patients. The VGG19-BN model delivered the best overall performance, with an accuracy of 93.76%, precision of 91.40%, specificity of 98.73%, and F1 score of 78.34%. The VGG16-BN model achieved the highest AUC value of 0.9787, while all models based on VGG architectures were superior to the others on the whole. The highest scores of six physicians in terms of accuracy, precision, specificity, and the F1 score were 74.00%, 75.00%, 88.33%, and 61.76%, respectively. The performance of each model of facial recognition was superior to that of the best physician on all metrics.ConclusionModels of computer-assisted facial recognition can improve the rate of diagnosis of NS. The models based on VGG19-BN and VGG16-BN can play an important role in diagnosing NS in clinical practice.

Automatic extraction of facial median sagittal plane for patients with asymmetry based on the EDMA alignment algorithm.

We aimed to establish a novel method for automatically constructing three-dimensional (3D) median sagittal plane (MSP) for mandibular deviation patients, which can increase the efficiency of aesthetic evaluating treatment progress. We developed a Euclidean weighted Procrustes analysis (EWPA) algorithm for extracting 3D facial MSP based on the Euclidean distance matrix analysis, automatically assigning weight to facial anatomical landmarks. Forty patients with mandibular deviation were recruited, and the Procrustes analysis (PA) algorithm based on the original mirror alignment and EWPA algorithm developed in this study were used to construct the MSP of each facial model of the patient as experimental groups 1 and 2, respectively. The expert-defined regional iterative closest point algorithm was used to construct the MSP as the reference group. The angle errors of the two experimental groups were compared to those of the reference group to evaluate their clinical suitability. The angle errors of the MSP constructed by the two EWPA and PA algorithms for the 40 patients were 1.39 ± 0.85°, 1.39 ± 0.78°, and 1.91 ± 0.80°, respectively. The two EWPA algorithms performed best in patients with moderate facial asymmetry, and in patients with severe facial asymmetry, the angle error was below 2°, which was a significant improvement over the PA algorithm. The clinical application of the EWPA algorithm based on 3D facial morphological analysis for constructing a 3D facial MSP for patients with mandibular deviated facial asymmetry deformity showed a significant improvement over the conventional PA algorithm and achieved the effect of a dental clinical expert-level diagnostic strategy.

Song Recommendation System Based on Facial Emotion

This research is aiming to enhance the user experience in music consumption by incorporating real-time facial emotion analysis. Emotions play a fundamental role in shaping individual preferences, and leveraging facial expressions as a means of understanding user’s emotional states can significantly contribute to personalized music recommendations. Our proposed system begins by capturing real-time facial expressions using a webcam or analyzing static images. These facial expressions are then processed through a CNN-based emotion recognition model trained to classify emotions such as happiness, sadness, anger, and more. The CNN model extracts high-level features from facial images, enabling accurate emotion recognition. Using the detected emotional state as input, our system employs a recommendation algorithm tailored to the user's current emotional state to suggest relevant music or videos from YouTube.

Multitask Learning Strategy with Pseudo-Labeling: Face Recognition, Facial Landmark Detection, and Head Pose Estimation

Most facial analysis methods perform well in standardized testing but not in real-world testing. The main reason is that training models cannot easily learn various human features and background noise, especially for facial landmark detection and head pose estimation tasks with limited and noisy training datasets. To alleviate the gap between standardized and real-world testing, we propose a pseudo-labeling technique using a face recognition dataset consisting of various people and background noise. The use of our pseudo-labeled training dataset can help to overcome the lack of diversity among the people in the dataset. Our integrated framework is constructed using complementary multitask learning methods to extract robust features for each task. Furthermore, introducing pseudo-labeling and multitask learning improves the face recognition performance by enabling the learning of pose-invariant features. Our method achieves state-of-the-art (SOTA) or near-SOTA performance on the AFLW2000-3D and BIWI datasets for facial landmark detection and head pose estimation, with competitive face verification performance on the IJB-C test dataset for face recognition. We demonstrate this through a novel testing methodology that categorizes cases as soft, medium, and hard based on the pose values of IJB-C. The proposed method achieves stable performance even when the dataset lacks diverse face identifications.

Enhancing facial geometry analysis by DeepFaceLandmark leveraging ResNet101 and transfer learning

Enhancing facial geometry analysis by DeepFaceLandmark leveraging ResNet101 and transfer learning