Facial Points Research Articles

Uncertain and biased facial expression recognition based on depthwise separable convolutional neural network with embedded attention mechanism

Datasets play a crucial role in the development of facial expression recognition (FER), but most of the datasets suffer from obvious uncertainties and biases caused by different cultures and collection conditions. To look deeper into these issues, this paper first conducts two sets of experiments, face detection, and facial expression (FE) classification. They are based on three datasets (CK+, FER2013, and RAF–DB), which are collected from lab and wild environments. This paper proposes a network, depthwise separable convolutional neural network (CNN) with an embedded attention mechanism (DSA–CNN) for expression recognition. First, at the preprocessing stage, we obtain the maximum expression range clipping, which is calculated from 81 facial landmark points to filter nonface interferences. Then, we use DSA–CNN, which is based on a coordinate squeeze-and-excitation (CSE) attention for feature extraction. Finally, to further deal with imbalanced class biases and uncertainties issues, this paper proposes a class-weighted cross-entropy loss (CCE-loss) to alleviate the imbalance among seven emotional classes. Then, we combine CCE-loss with ranking regularization loss (RR-loss) and self-importance weighting cross-entropy loss (SCE-loss) at label amend stage, to jointly guide the training of the network. Extensive experiments on three FER datasets demonstrate that our proposed method outperforms most of the state-of-the-art methods eventually.

Objective evaluation of mechanical expressiveness in android and human faces

ABSTRACT Extensive research has been conducted to realize highly expressive android faces. In contrast, not much is being discussed which android faces are mechanically better and how advanced each android is because a fair and meaningful standard index of expressiveness does not exist. Therefore, in this study, we propose a numerical index and three types of visualization methods for facial deformation expressiveness to investigate the mechanical performance of android faces compared to humans. We used a data set of three-dimensional displacement vectors of over 100 facial points on the faces of two androids and three human males to calculate the expressiveness that reflects how diversely each facial point could move. As a result, we found that the expressiveness of the androids was significantly lower than humans, especially in the lower faces. The expressiveness of the androids was approximately only 20 that of humans even in the most expressive axis and two orders of magnitude lower than humans in three-dimensional evaluation. Thus, our method revealed the lackness of mechanical performance in recently developed androids to move facial skin widely. This method enabled us to reasonably discuss future directions to realize next generation androids with expressiveness similar to that in humans.

A new instrument to clinically evaluate the anteroposterior relationship of the maxillary central incisors to the forehead

To evaluate the reliability of a new instrument to clinically measure the anteroposterior relationship of the maxillary central incisors to the forehead. An analytical cross-sectional study using a sample of 84 (67% female and 33% male) young adult participants was conducted. Measurements were performed according to Element II of the Andrews' Six Elements of Orofacial Harmony by assessing the horizontal distance between the facial axis point of the maxillary central incisor and the forehead's anterior limit line. The instrument tested was a transparent plastic template having a leveling meter, reference lines, and an embedded millimeter ruler. Two orthodontists evaluated the sample twice with a minimum of a two-week interval between evaluations. The reliability was evaluated using the intraclass correlation coefficient. Repeated-measures ANOVA was also performed. Good to excellent intra- and interexaminer agreement, with intraclass correlation coefficient values of 0.874 (95% confidence interval 0.800-0.921; P < 0.001) and 0.876 (95% confidence interval 0.739-0.948; P < 0.001), respectively, were calculated. There were no statistically significant differences between all the repeated measurements assessed (P=0.820). The instrument showed good to excellent reliability for determining the anteroposterior relationship of the maxillary central incisors to the forehead and could be used to help orthodontists and maxillofacial surgeons indirectly evaluate the anteroposterior position of the maxilla.

E-faceatlasAR: extend atlas of facial acupuncture points with auricular maps in augmented reality for self-acupressure

E-faceatlasAR: extend atlas of facial acupuncture points with auricular maps in augmented reality for self-acupressure

Accuracy of 3-dimensional printed bracket transfer tray using an in-office indirect bonding system

In this study, we aimed to evaluate and compare the bracket positioning accuracy of the indirect bonding (IDB) transfer tray fabricated in-clinic using the tray printing (TP) and marker-model printing methods (MP). The TP group was further divided into 2 groups (single-tray printing [STP] and multiple-tray printing [MTP]) depending on the presence of a tray split created using the 3-dimensional (3D) software. Five duplicated plaster models were used for each of the 3 experimental groups, and a total of 180 artificial teeth, except the second molar, were evaluated in the experiment. The dental model was scanned using a model scanner (E3; 3Shape Dental Systems, Copenhagen, Denmark). Virtual brackets were placed on facial axis points, and the IDB trays were designed and fabricated using a 3D printer (VIDA; EnvisionTEC, Mich). The accuracy of bracket positioning was evaluated by comparing the planned bracket positions and the actual bracket positions using 3D analysis on inspection software. The main effects and first-order interaction effects were analyzed together by analysis for the analysis of variance. The mean distance and height errors were significantly lower in the STP group than those in the MP and MTP groups (P<0.05). The mean distance error was 0.06 mm in the STP group and 0.09 mm in the MP and MTP groups. The mean height error was 0.10 mm in the STP group and 0.15 mm and 0.18 mm in MP and MTP groups, respectively. However, no significant differences were observed in the angular errors among the 3 groups. The in-office-fabricated IDB system with computer-aided design and 3D printer is clinically applicable after considering the linear and angular errors. We recommend IDB trays fabricated using the STP method owing to the lower frequency of bracket positioning errors and ease of fabrication.

Facial Emotion Recognition with Music Recommendation

Facial emotion recognition is advancement in computer vision and machine literacy and with the help of this computing technology it's easy to identify mortal emotion through images also. In this paper we propose fashion call facial emotion recognition with music recommendation using Convolutional Neural Network (CNN). The FER is grounded on three corridor. The first part removes the background from the picture, the alternate part concentrates and maps the facial point vector birth, and the third part recommends music grounded on prognosticated emotion. So, to train the images online database is taken from the Kaggle and consequently the feelings are labelled with 96 of delicacy. Further, grounded on emotion vaticination music or audio song will be recommended from the database.

Performance Comparison of Feature Face Detection Algorithm on The Embedded Platform

The intensity of light will greatly affect every process carried out in image processing, especially facial images. It is important to analyze how the performance of each face detection method when tested at several lighting levels. In face detection, various methods can be used and have been tested. The FLP method automates the identification of the location of facial points. The Fisherface method reduces the dimensions obtained from PCA calculations. The LBPH method converts the texture of a face image into a binary value, while the WNNs method uses RAM to process image data, using the WiSARD architecture. This study proposes a technique for testing the effect of light on the performance of face detection methods, on an embedded platform. The highest accuracy was achieved by the LBPH and WNNs methods with an accuracy value of 98% at a lighting level of 400 lx. Meanwhile, at the lowest lighting level of 175 lx, all methods have a fairly good level of accuracy, which is between 75% to 83%.

How Do You Treat Trigeminal Neuralgia in Your Practice?

How Do You Treat Trigeminal Neuralgia in Your Practice?

Robust facial expression recognition system in higher poses

Facial expression recognition (FER) has numerous applications in computer security, neuroscience, psychology, and engineering. Owing to its non-intrusiveness, it is considered a useful technology for combating crime. However, FER is plagued with several challenges, the most serious of which is its poor prediction accuracy in severe head poses. The aim of this study, therefore, is to improve the recognition accuracy in severe head poses by proposing a robust 3D head-tracking algorithm based on an ellipsoidal model, advanced ensemble of AdaBoost, and saturated vector machine (SVM). The FER features are tracked from one frame to the next using the ellipsoidal tracking model, and the visible expressive facial key points are extracted using Gabor filters. The ensemble algorithm (Ada-AdaSVM) is then used for feature selection and classification. The proposed technique is evaluated using the Bosphorus, BU-3DFE, MMI, CK + , and BP4D-Spontaneous facial expression databases. The overall performance is outstanding.

Spatiotemporal CNN with Pyramid Bottleneck Blocks: Application to eye blinking detection

Eye blink detection is a challenging problem that many researchers are working on because it has the potential to solve many facial analysis tasks, such as face anti-spoofing, driver drowsiness detection, and some health disorders. There have been few attempts to detect blinking in the wild scenario, while most of the work has been done under controlled conditions. Moreover, current learning approaches are designed to process sequences that contain only a single blink ignoring the case of the presence of multiple eye blinks. In this work, we propose a fast framework for eye blink detection and eye blink verification that can effectively extract multiple blinks from image sequences considering several challenges such as lighting changes, variety of poses, and change in appearance. The proposed framework employs fast landmarks detector to extract multiple facial key points including the ones that identify the eye regions. Then, an SVD-based method is proposed to extract the potential eye blinks in a moving time window that is updated with new images every second. Finally, the detected blink candidates are verified using a 2D Pyramidal Bottleneck Block Network (PBBN). We also propose an alternative approach that uses a sequence of frames instead of an image as input and employs a continuous 3D PBBN that follows most of the state-of-the-art approaches schemes. Experimental results show the better performance of the proposed approach compared to the state-of-the-art approaches.

Facial Expression Analysis and Estimation Based on Facial Salient Points and Action Unit (AUs)

Humans use their facial expressions as one of the most effective, quick, and natural ways to convey their feelings and intentions to others. In this research, presents the analyses of human facial structure along with its components using Facial Action Units (AUs) and Geometric structures for identifying human facial expressions. The approach considers facial components such as Nose, Mouth, eyes and eye brows for FER. Nostril contours such as left lower tip, right lower tip, and centre tip are considered as salient points of Nose. Various salient points for Mouth are extracted from the left and right end point, upper and lower lip mid points along with curve. These salient points are extracted for all facial expression of the same subject considering neutral face as reference. The Geometric structure for neutral face is mapped along with other facial expression faces. The deformation is estimated using the Euclidean distance. The classification algorithms such as LibSVM, MLP, RF has achieved classification accuracy of 86.56% on an average. The findings of the experiments show that the extraction of picture characteristics is more efficient in terms of computing and gives promising outcomes.

Accuracy of dental arch form in customized fixed labial orthodontic appliances

This study evaluated arch form accuracy with or without premolar extraction in customized fixed labial orthodontic appliance treatment. Setup and posttreatment digital models of 27 samples (15 extractions and 12 nonextractions) were selected and superimposed by best-fit surface-based registration in both the maxilla and the mandible. The facial axis points were identified and converted into Cartesian coordinates. A sixth-order polynomial equation was used to fit dental arches. Arch discrepancies (the mean distance between 2 arch forms) and similarities were compared between extraction and nonextraction groups, maxilla and mandible, and anterior and posterior arches. The arch discrepancy between extraction and nonextraction groups showed no statistically significant difference, but a statistically significant difference in arch similarity was found in the mandible. There were statistically significant differences between anterior and posterior arch discrepancies in the extraction (mandible) and the nonextraction (maxilla and mandible) groups. However, no statistically significant correlation was shown between anterior and posterior arch discrepancies. The arch similarities were 96.18% and 97.38% in the maxilla and 96.01% and 97.49% in the mandible between extraction and nonextraction groups. Arch form discrepancies and similarities showed a moderate correlation but no statistically significant differences between the maxilla and the mandible. In customized fixed labial orthodontic appliance treatment, arch form setup can be accurately achieved with and without premolar extraction. Anterior arch form acquires fewer discrepancies than the posterior arch, and overcorrection should be added to the end of the customized archwire to reduce posterior arch discrepancies. The discrepancy of the maxillary and mandibular arches is interrelated, and adjustments should be made on both maxillary and mandibular archwires to correct single-jaw transverse malposition.

Multi-feature based automatic facial expression recognition using deep convolutional neural network

&lt;p&gt;&lt;span&gt;Deep multi-task learning is one of the most challenging research topics widely explored in the field of recognition of facial expression. Most deep learning models rely on the class labels details by eliminating the local information of the sample data which deteriorates the performance of the recognition system. This paper proposes multi-feature-based deep convolutional neural networks (D-CNN) that identify the facial expression of the human face. To enhance the accuracy of recognition systems, the multi-feature learning model is employed in this study. The input images are preprocessed and enhanced via three filtering methods i.e., Gaussian, Wiener, and adaptive mean filtering. The preprocessed image is then segmented using a face detection algorithm. The detected face is further applied with local binary pattern (LBP) that extracts the facial points of each facial expression. These are then fed into the D-CNN that effectively recognizes the facial expression using the features of facial points. The proposed D-CNN is implemented, and the results are compared to the existing support vector machine (SVM). The analysis of deep features helps to extract the local information from the data without incurring a higher computational effort.&lt;/span&gt;&lt;/p&gt;

Facial landmark-guided surface matching for image-to-patient registration with an RGB-D camera.

Fiducial marker-based image-to-patient registration is the most common way in image-guided neurosurgery, which is labour-intensive, time consuming, invasive and error prone. We proposed a method of facial landmark-guided surface matching for image-to-patient registration using an RGB-D camera. Five facial landmarks are localised from preoperative magnetic resonance (MR) images using deep learning and RGB image using Adaboost with multi-scale block local binary patterns, respectively. The registration of two facial surface point clouds derived from MR images and RGB-D data is initialised by aligning these five landmarks and further refined by weighted iterative closest point algorithm. Phantom experiment results show the target registration error is less than 3mm when the distance from the camera to the phantom is less than 1000mm. The registration takes less than 10s. The proposed method is comparable to the state-of-the-arts in terms of the accuracy yet more time-saving and non-invasive.

Attractive Versus Golden Ratios: Formula of Utopian Beauty through Comparison of Facial Proportions among Worldwide Celebrities and Average People.

Beauty criteria change according to trends, cultures, and expectations. So, the golden ratios are not enough to understand the actual beauty concept. It is aimed to create a method and formula to define beauty in terms of different facial measurements and ratios as variables. Frontal view photographs of 50 celebrities and 50 volunteers (female [ n = 70] and male [ n = 30]) were obtained. Thirty-three specific facial points were used. Fifty-nine measurements of distances between those points and ratios of those distances are used as statistical variables. Mean values of those variables for celebrities are compared with all participants.Then, ''total beauty score'' using statically significant differences is defined. Points according to the importance level of interested value are assigned; 1 point for a significance of 0.001 ≤ P < 0.05 and 2 points for P < 0.001. Possible independent variables for the prediction of Utopian Beauty Score were determined using linear regression analysis. Finally, 10 variables (6 measurements and 4 ratios) are used to determine a formula for beauty. In celebrities, the distance between head apex and hair border, lower lip vertical length, the distance between brow medial borders and forehead vertical length were higher than the average population ( P < 0.001, 0.023, 0.034, 0.001, respectively). However, the width of columella, a length between bilateral ala nasi, horizontal face, and vertical nose length, and the distance between brow apexeyelash line were shorter ( P = 0.005, < 0.001, < 0.001, < 0.001, < 0.001, respectively). Some study variables had severely significant correlations with the total beauty point (M29, r = -0.744, P < 0.001; R19, r =-0.745, P < 0.001; and R30, r = -0.735, P < 0.001). The linear regression formula for Utopian Beauty Score was determined according to statistically significant variables as y = 86.5 - 3.6 M5 + 8.1 M14 + 11.1 M20 - 6.4 M25-8.7 M29-10.3 M30-15.6 R19 + 9.3 R20 + 16.4 R25 + 18.3 R26. A large forehead, small nose with lifted tip, plump lip, brows with lateral apex, well-spaced eyes, and ovoid, elliptical face lines are seem to be more attractive. Additionally, a method to create a formula for beauty with variables of different measurements and ratios of facial points are determined.

How to choose the size of facial areas of interest in interactive eye tracking.

Advances in eye tracking technology have enabled the development of interactive experimental setups to study social attention. Since these setups differ substantially from the eye tracker manufacturer's test conditions, validation is essential with regard to the quality of gaze data and other factors potentially threatening the validity of this signal. In this study, we evaluated the impact of accuracy and areas of interest (AOIs) size on the classification of simulated gaze (fixation) data. We defined AOIs of different sizes using the Limited-Radius Voronoi-Tessellation (LRVT) method, and simulated gaze data for facial target points with varying accuracy. As hypothesized, we found that accuracy and AOI size had strong effects on gaze classification. In addition, these effects were not independent and differed in falsely classified gaze inside AOIs (Type I errors; false alarms) and falsely classified gaze outside the predefined AOIs (Type II errors; misses). Our results indicate that smaller AOIs generally minimize false classifications as long as accuracy is good enough. For studies with lower accuracy, Type II errors can still be compensated to some extent by using larger AOIs, but at the cost of more probable Type I errors. Proper estimation of accuracy is therefore essential for making informed decisions regarding the size of AOIs in eye tracking research.

Systems Thinking

Systems Thinking

Automatic Real-Time Medical Mask Detection Using Deep Learning to Fight COVID-19

The COVID-19 pandemic is a virus that has disastrous effects on human lives globally; still spreading like wildfire causing huge losses to humanity and economies. There is a need to follow few constraints like social distancing norms, personal hygiene, and masking up to effectively control the virus spread. The proposal is to detect the face frame and confirm the faces are properly covered with masks. By applying the concepts of Deep learning, the results obtained for mask detection are found to be effective. The system is trained using 4500 images to accurately judge and justify its accuracy. The aim is to develop an algorithm to automatically detect a mask, but the approach does not facilitate the percentage of improper usage. Accuracy levels are as low as 50% if the mask is improperly covered and an alert is raised for improper placement. It can be used at traffic places and social gatherings for the prevention of virus transmission. It works by first locating the region of interest by creating a frame boundary, then facial points are picked up to detect and concentrate on specific features. The training on the input images is performed using different epochs until the artificial face mask detection dataset is created. The system is implemented using TensorFlow with OpenCV and Python using a Jupyter Notebook simulation environment. The training dataset used is collected from a set of diverse open-source datasets with filtered images available at Kaggle Medical Mask Dataset by Mikolaj Witkowski, Kera, and Prajna Bhandary. To simulate MobilNetV2 classifier is used to load and pre-process the image dataset for building a fully connected head. The objective is to assess the accuracy of the identification, measuring the efficiency and effectiveness of algorithms for precision, recall, and F1 score.

Key facial points recognition using ResNet

Key facial points recognition is a major research topic in the domain of image processing. In this process various important facial points such as nose tip, eye center etc are detected using machine learning and AI. Key facial points recognition can help for the face emotion recognition. Although convolutional neural networks (CNN) can be used for face recognition but the CNN suffers from the vanishing gradient problem. While training the CNN model using back propagation algorithm the vanishing gradient problem occurs. Therefore the models stops training as the gradient become vanishingly small. To solve this problem a ResNet model has been implemented in this paper. The ResNet model solves the vanishing gradient problem with the help of an additional skip connection. And therefore a deep neural network can be constructed using the ResNet. The proposed ResNet model identifies the key facial points with high accuracy. After identifying the key facial points one can easily identify the emotion class of the face image. The detected key facial points in this experiment are left eye center, right eye center, left eye inner corner, left eye outer corner, right eye inner corner, nose tip, mouth left corner, mouth right corner, mouth right corner, mouth center top lip, mouth center bottom lip. After training the proposed ResNet model for 100 epochs it successfully detected the key facial points with 77.70 percent accuracy.

Intra-socket application of Hyaluronic acid reduces pain and swelling following wisdom teeth removal

Background:Hyaluronic acid (HA) has the potential to promote wound healing. Pain and swelling with trismus are common sequalae post wisdom teeth removal. This study aims to investigate the effectiveness of intra-socket HA solution to reduce these uncomfortable post-operative events.Materials and methods:30 patients underwent bilateral extractions of mandibular wisdom teeth for this study. Intra-socket application of 0.7 ml 20 mg/2 ml HA solution (Hyalgun) with Gel foam as a scaffold in study site versus Gel foam only on control site was conducted via a split mouth study design. Data collection of five facial reference points for swelling and maximum mouth opening was recorded during the pre-operative period and post-operative 2nd and 7th day. The VAS pain score at post-op 1st, 2nd and 3rd day and the number of analgesics for the 7-day post-operative duration were evaluated.Results:The HA group demonstrated statistically significant less swelling, trismus and analgesia consumption on the 2 and 7 days after surgery. VAS scores on day 1, 2 and 3 after surgery (P = 0.05) were significantly less in the HA group compared to the control group.Conclusion:The application of intra-socket HA has a positive effect for reducing postoperative pain and swelling with trismus after the lower third molar intervention (LTMI).