Part Of Face Research Articles

Developing a Machine Learning-Based Algorithm for Facial Expression Recognition

With the rise of artificial intelligence, machine learning (ML) is increasingly integrated into daily life. Facial expressions, lasting about 1/20th of a second and difficult to conceal, convey nuanced emotions beyond words. They are categorized as macro expressions, displayed under normal circumstances, and micro expressions, fleeting and subconscious. However, recognizing expressions in photos is challenging due to varied backgrounds, appearances, age, and race, impacting accuracy. Addressing this, our research focuses on a machine learning-based facial expression recognition algorithm. Our experimental results, using CASIA and NTERFACE databases, demonstrate that ML outperforms KNN and ANN with recognition rates 41.54% and 26.94% higher, respectively. By adaptively learning global and local features from complete expression images and face parts, our algorithm offers practical and significant advancements in expression recognition.

The Crucial Role of the Eyes in Predicting Facial Attractiveness from Parts

Perceiving facial attractiveness based on partial information is a common experience in social interactions, where we often view only parts of faces. However, our understanding of predictions of the whole face from its parts remains limited. This study investigated this by dividing faces into the top and bottom halves (Study 1) and the forehead, eyes, nose, and mouth (Study 2). We compared how attractiveness was predicted from these parts versus when the whole face was fully observed. In the findings, overestimations occurred when predicting from the eyes (or the top half). Similarly, predictions from the nose resulted in overestimations, though to a lesser extent. In contrast, no difference was found between the predicted and observed attractiveness of the whole face when predicting from the mouth (or the bottom half). Further analysis revealed that the eyes were the most significant predictor of facial attractiveness across all facial features. Interestingly, higher attractiveness of the eyes was associated with a greater extent of overestimations in predictions from the eyes, whereas predictions from other parts exhibited a reduced extent of overestimations. These findings implied that predictions were influenced by the eyes, underscoring the critical role the eyes play in the process of predicting facial attractiveness.

Endovascular Preparation With Innovative Custom-Made Stent-Graft Before Kidney Transplantation: The Solution for Patients With Hostile Iliac Calcification.

The increasing age of patients with end-stage renal disease raises the issue of hostile arterial access for transplantation, with technical difficulties associated with clamping and suturing the iliac artery. Some of these patients - who theoretically represent those who would benefit the most from transplantation in terms of mortality - are contraindicated because of anatomical and medical issues. In this context, a specific endovascular device called EndoPreKiT (Endovascular Preparation for Kidney Transplantation) has been designed, enabling arterial access for transplantation via a mini-invasive procedure. It consists of a woven Dacron supported by self-expanding nitinol rings, ensuring anchorage and allowing arterial clamping. The middle part of the anterior face of the device is stentless, enabling the anastomosis directly onto the Dacron once the calcified artery wall has been removed. After a cadaveric study validating its technical feasibility, such device was successfully implanted in 10 patients considered unfit for transplantation due to severe wall calcification. Two of them have been successfully transplanted with excellent outcomes after 13 and 3months of follow-up. EndoPreKiT device may be a significant breakthrough in transplant surgery, that could expand the horizon of eligibility to include even the most fragile patients with challenging arterial access.

Impact of partial occlusion of the face on multisensory emotion perception: Comparison of pre- and post-COVID-19 pandemic.

We perceive and understand others' emotional states from multisensory information such as facial expressions and vocal cues. However, such cues are not always available or clear. Can partial loss of visual cues affect multisensory emotion perception? In addition, the COVID-19 pandemic has led to the widespread use of face masks, which can reduce some facial cues used in emotion perception. Thus, can frequent exposure to masked faces affect emotion perception? We conducted an emotion perception task using audio-visual stimuli that partially occluded the speaker's face. Participants were simultaneously shown a face and voice that expressed either congruent or incongruent emotions and judged whether the person was happy or angry. The stimuli included videos in which the eyes or mouth were partially covered and where the whole face was visible. Our findings showed that, when facial cues were partially occluded, participants relied more on vocal cues for emotion recognition. Moreover, when the mouth was covered, participants relied less on vocal cues after the pandemic compared to before. These findings indicate that partial face masking and prolonged exposure to masked faces can affect multisensory emotion perception. In unimodal emotion perception from only facial cues, accuracy also improved after the pandemic compared to before for faces with the mouth occluded. Therefore, changes in the reliance on vocal cues in multisensory emotion perception during the pandemic period could be explained by improved facial emotion perception from the eye region.

‘Richt honorabill’ ladies: Noble Power and Aristocratic Women in Sixteenth-Century Scotland

This essay began as a talk given in October 2023 to inaugurate an annual public lecture in honour of Jenny Wormald. It takes its inspiration from her work on noble power, re-examining how noblewomen in sixteenth-century Scotland employed that power. Visual and cultural evidence has been prioritised in the essay which is organised around two settings: the royal court and noble country seats. As the Stirling Heads show, noblewomen formed part of the public face of the Scottish court. They also exploited their right of access to the monarch to further their own and family interests. The arrival of renaissance country houses expanded noblewomen's central duty of running their residences, sometimes including managing the family estates. ‘Holding hous’ required the provision of hospitality and of the new country-house culture. Noblewomen formed a vital communication system linking the regions to the capital and court. Their informal networks helped negotiate the inter-relationships of birth and marital kindreds, of friendships and alliances. The female contribution to the success of noble families suggests a broader definition of Scottish noble power is required. Further exploration might also place that experience within the European context of the development of aristocratic elites.

Face cells encode object parts more than facial configuration of illusory faces

Humans perceive illusory faces in everyday objects with a face-like configuration, an illusion known as face pareidolia. Face-selective regions in humans and monkeys, believed to underlie face perception, have been shown to respond to face pareidolia images. Here, we investigated whether pareidolia selectivity in macaque inferotemporal cortex is explained by the face-like configuration that drives the human perception of illusory faces. We found that face cells responded selectively to pareidolia images. This selectivity did not correlate with human faceness ratings and did not require the face-like configuration. Instead, it was driven primarily by the “eye” parts of the illusory face, which are simply object parts when viewed in isolation. In contrast, human perceptual pareidolia relied primarily on the global configuration and could not be explained by “eye” parts. Our results indicate that face-cells encode local, generic features of illusory faces, in misalignment with human visual perception, which requires holistic information.

The skull of Hapalodectes (Hapalodectidae, Mesonychia) from the Paleocene of Mongolia.

The incomplete skull of Hapalodectes (Hapalodectidae, Mesonychia) is described from the Upper Paleocene Naran Member of the Naran Bulak Formation of the Khaychin-Ula 2 locality in the Bugin Tsav Basin in the south of Mongolia. This is the first discovery of the skull of a Paleocene member of the genus. The skull corresponds in size and dental occlusion to the species H. dux Lopatin, 2001, previously known only from a lower jaw from the Upper Paleocene Zhigden Member of the Naran Bulak Formation of the Tsagan-Khushu locality in the Nemegt Basin. In the structure of the facial part of the skull, H. dux exhibits features shared with Mesonychidae, namely the significant participation of the lacrimal and jugal in the preorbital region. The upper molars of H. dux were found to have distinct conules (a first for Hapalodectidae). The evolutionary changes in the structure of the upper molars of Hapalodectes (disappearance of the conules, decrease of the metacone, and narrowing of the lingual lobe) are determined. They were correlated with the transformations of the lower molars (reduction of the metaconid, protocristid, and entoconid, transverse compression of the trigonid and talonid) and together with them were aimed at enhancing the shearing action of the longitudinal trenchant blades of the posterior part of the dentitions.

Investigation of the specific charge variations in determining the optimal methods for drilling and blasting of tunnels

In this research, in order to compare and select the optimum patterns, two parameters of specific charge and specific drilling are used. Two explosive substances, including ammonium nitrate and fuel oil (ANFO) and Amulet, are used to charge the holes located on the cross-section of the tunnels in order to advance by 3 meters. Also, the cross-section of the selected tunnels was 19, 28, 32, 40, 48, 50, 65, and 76 square meters with horseshoe, circle and D-shaped. The diameter of the explosive hole is 48 mm for the Norwegian method and 51 mm for the two energy balance models and the Swedish method to compare them. A total of 160 designs are prepared for three types of limestone, sandstone and marl. The results obtained for two parameters of specific charge and special drilling for the selected rock material are completely opposite. Also, in the Swedish methods, according to the Amulet, the values of the calculated parameters are close to the energy balance model. Finally, according to the geomechanical parameters of the rock, for the Swedish method and the energy balance model in charging with amulet explosives, the energy balance model and the Holemberg-Persson method are proposed in the optimal design for the tunnel drilling and blasting pattern. In addition, for the Norwegian method and the energy balance model in charging the holes with ANFO, the energy balance model is prior to the Norwegian method. The amount of stemming of holes in different parts of the tunnel face is less for the Swedish method than the Norwegian and energy balance models.

New Extracted Features to Recognize Faces Effected by Occlusions and Common Variations

Face recognition from non-identical face photos is a prominent area of research in pattern recognition and computer vision. Existing face recognition systems struggle with diverse changes like lighting conditions, expressions, and facial occlusions. This paper proposes a new Face Recognition (FR) approach that combines the Elastic Bunch Graph Matching (EBGM) approach with the greedy algorithm to automatically identify face landmarks. The proposed approach independently selects each optimal landmark of face image from different corresponding face images where the corresponding landmark of corresponding face image which achieves the best similarity is used rather than using one or at most two corresponding face images and computing the average between both. The locations of corresponding landmarks can be displaced to achieve maximum similarity with optimal landmarks. This proposed approach demonstrates improved recognition performance compared to contemporary face recognition methods. It effectively handles changing ratios of face parts and can recognize faces even with increasing occlusion sizes.

Mechanisms of functional asymmetry of the face formation in patients with dysfunctional condition of temporomandibular joint

Objective. To give a detailed description of clinical manifestations, diagnosis and treatment of functional asymmetry of the face in patients with dysfunctional condition of temporomandibular joint (TMJ). Materials and methods. A complete examination of 34 (25 %) patients with symptomatic facial asymmetry of the 3rd type according to H.S. Hwang classification was carried out. The group consisted of 5 males (14,71 %) and 29 females (85,29 %). Screening protocol included standard dental care, cone beam computer tomography of both jaws and TMJ, MRI of TMJ with functional load. Results. 24 patients (70.59 %) completed complex dental treatment; 10 (29.41 %) people stopped treatment for various reasons and were excluded from further study. Complete restoration of facial symmetry was achieved in 21 people (87.5 %), partial restoration in 3 (12.5 %) patients. The height of the lower facial part and the synchrony of occlusal contacts were restored in all patients. Noise in the joint area persisted in 3 (8.8 %) cases, which is associated with arthrosis in the articular heads. In patients with unreducible dislocations, the intensity of the clicks decreased, and there was no pain at the end of treatment. It should be noted that all patients of this group refused subsequent surgical treatment. Conclusion. Functional asymmetry of the face results from multifactorial impact on the human organism and is multimorbid by nature. This condition is accompanied by complicated dental pathology, such as increased hard tooth tissues attrition, functional overload with traumatic nodes formation, disturbance of muscles and TMJ functions. It requires following diagnostic and treatment algorithm under the supervision of internal medicine specialists to correct sleep, postural stability, impaired psychoneurological status.

Gender dimorphism of the relationships between the cranial and facial skull sections in mature human individuals

Background. To date, the range of individual anatomical variability of the cranial profile type in mature adults, depending on gender and skull structure type has not yet been adequately established. Aim. To study gender-specific ratios of the cranial and facial parts of the skull in mature adults. Materials and Methods. The materials of our study consisted of 115 skulls of mature adults of both sexes, including 35 dry bone specimens of whole or fragmented skulls. For each parameter, the following metrics were calculated: The following parameters were used in the statistical analysis: (arithmetic mean), σ (standard deviation), and m (standard error of the mean). Results. It was found that in men the skull length is =181.36 with σ=3.31 and m=0.46, while in women it does not exceed =176.57 with σ=3.28 and m=0.47; the skull width in men reaches =141.17 with σ=2.62 and m=0.34, and in women, it does not exceed =134.94 with σ=2.50 and m=0.33; the height parameter of the skull in men also reaches maximum values at the level of =141.38 with σ=3.75 and m=0.52, while in women, it is somewhat lower at =137.24 with σ=3.15 and m=0.55. In mature men, the upper facial width is =94.17 with σ=3.87 and m=0.50, while in women, this parameter averages =90.71 with σ=4.05 and m=0.54; the middle width in men reaches =128.93 with σ=1.53 and m=0.20, while in women, it does not exceed =122.49 with σ=1.44 and m=0.19; the maximum indicators of angular width are established in men, averaging =99.64 with σ=4.15 and m=0.54, while in women, they do not exceed =92.27 with σ=3.78 and m=0.51. Conclusions. The study identified significant gender-specific differences in the cranial and facial dimensions of middle-aged individuals. Men demonstrated larger average values for skull length, width, and height, as well as upper face width, bizygomatic width, and angular width, compared to women. These findings highlight distinct morphological characteristics between genders. Keywords: computed tomography, cranial index, face width, face length, face height.

Robust Face Recognition System Based on Deep Facial Feature Extraction and Machine Learning

Facial Recognition is a common challenge in research for security applications. Wide-ranging applications of this technology include biometrics, security data, access to controlled locations, smart cards, and surveillance systems. A practical approach to identifying individuals depends on factors such as partial face occlusion, illumination, age, expression, makeup, and poses. These complex face recognition variables affect most face recognition systems. In this paper, we propose a deep transfer learning for facial features based on the pose and illumination variations of celebrities’ faces in a free environment for face recognition. The new approach that combines deep learning methods for feature extraction with machine learning classification methods. Also, studies the performance of the pre-trained model (VGGFace2 with ResNet-50 model weights) for feature extraction with a Multilayer Perceptron Classifier (MLP), Decision Tree, and Bootstrap Aggregation (Bagging) to perform classification. The convolution neural network has lately achieved excellent advancement in facial recognition. The outcome indicated that VGGFace2 with MLP obtained more precision and provided the best results (Accuracy, F-measure, Recall, and Precision). The results for all metric for the proposed model is 99%. The benchmark for this model is a dataset of 105 people’s faces.

Synthesis of Comic-Style Portraits Using Combination of CycleGAN and Pix2Pix

This paper proposes a system for converting face photos into portraits with specific charcoal sketch-like style. NPR (non-photorealistic rendering) for converting real photo images into anime-style images have been studied. Its promising application is the creation of users’ portraits on social networks for preventing leaks of personal information. An image-to-image transformation using GAN (Generative Adversarial Network), such as CycleGAN or Pix2Pix, is expected to be used for this goal. However, it is difficult to generate portraits with specific comic styles that satisfy two conditions at the same time: preserving photorealistic features and reproducing the comic styles. For example, replacing too many photorealistic features on the face with comic style will destroy users’ identity, such as eyes shape. To solve this problem, the proposed system combines CycleGAN and Pix2Pix. CycleGAN is used to generate paired examples for Pix2Pix and Pix2Pix learns photo-to-comic transformation. To further emphasize a comic style, this paper also proposes two extensions, which divides face images into several parts such as hair and eye parts. The quality of generated images are evaluated with questionnaire by 50 answerers. The results show that the images generated by the proposed system are highly evaluated than the images generated by CycleGAN in terms of preserving the features in a photo and reproducing a comic style. Although the segmentation tend to collapse the generated images, it can reproduce a comic style effectively, especially on facial parts.

Visual self-face and self-body recognition in a left-brain-damaged prosopagnosic patient.

The present case study describes the patient N.G., who reported prosopagnosia along with difficulty in recognising herself in the mirror following a left-sided temporo-occipital hemispheric stroke. The neuropsychological and experimental investigation revealed only a mild form of apperceptive prosopagnosia, without visual agnosia, primarily caused by an impaired visual processing of face-parts and body parts but not of full faces. Emotional expressions did not modulate her face processing. On the other hand, N.G. showed a marked impairment of visual self-recognition, as assessed with visual matching-to-sample tasks, both at the level of body-part and face-part processing and at a full-face level, featured by a deficit in the perceptual discrimination of her own face and body, as compared to the others' face and body. N.G.'s lesion mapping showed damage to the left inferior occipito-temporal cortex, affecting the inferior occipital gyrus and compromising long-range connections between the occipital/temporo-occipital areas and the anterior fronto-temporal areas. Overall, the present case report documents that visual processing of the person's own face may be selectively compromised by a left-sided hemispheric lesion disconnecting extra-striate body- and face-selective visual areas to self-representation regions. Moreover, others' (full) face processing may be preserved, as compared with the impaired ability to discriminate others' body and face parts.

Comparative evaluation of facial divine proportions in males and females with average growth pattern

Facial beauty of an individual depends on proportion and its relationship to various facial parts which varies in each individual. The values of measured proportion in beautiful faces are likely to approximate divine proportion and may show variation with ethinicity, sex, growth pattern. Hence, aim of the study was to evaluate and compare vertical and horizontal parameters of facial divine proportions in male and female subjects with average growth pattern. 30 subjects (18-30 yrs age) with average growth pattern (based on cephalometric measurements FMA, Yaxis, GoGn-SN and facial axis angle) were divided into two groups as Group I (Male subjects, n=15) ; Group II (female subjects, n=15). Standardized frontal facial photographs of all the subjects were obtained. Nine vertical and four horizontal parameters were measured on all the facial photographs using IC Measure software and were used to evaluate horizontal and vertical facial proportions and then compared to divine proportions considering 1.618 as 100%. Horizontal and Vertical facial proportion in males were higher than females. Group I showed that three of seven vertical facial proportions were close to divine proportion (1.618) whereas only one vertical facial proportion in Group II were close to divine proportions. Horizontal facial proportions in both the groups deviated more from divine proportion (1.618).There was difference between males and females for the vertical and horizontal facial proportions with values being larger in males. This should be considered as an accessory guideline in planning orthodontic or orthognathic treatment.

Perceived Ambiguity of Facial Expression -Contribution of the Combination of Facial Parts of Different Emotions-

Perceived Ambiguity of Facial Expression -Contribution of the Combination of Facial Parts of Different Emotions-

Text2Face: Text-Based Face Generation With Geometry and Appearance Control.

Recent years have witnessed the emergence of various techniques proposed for text-based human face generation and manipulation. Such methods, targeting bridging the semantic gap between text and visual contents, provide users with a deft hand to turn ideas into visuals via text interface and enable more diversified multimedia applications. However, due to the flexibility of linguistic expressiveness, the mapping from sentences to desired facial images is clearly many-to-many, causing ambiguities during text-to-face generation. To alleviate these ambiguities, we introduce a local-to-global framework with two graph neural networks (one for geometry and the other for appearance) embedded to model the inter-dependency among facial parts. This is based upon our key observation that the geometry and appearance attributes among different facial components are not mutually independent, i.e., the combinations of part-level facial features are not arbitrary and thus do not conform to a uniform distribution. By learning from the dataset distribution and enabling recommendations given partial descriptions of human faces, these networks are highly suitable for our text-to-face task. Our method is capable of generating high-quality attribute-conditioned facial images from text. Extensive experiments have confirmed the superiority and usability of our method over the prior art.

Romantic bias in judging the attractiveness of faces wearing masks.

The spread of COVID-19 has drastically increased the number of people wearing masks in public areas and the opportunities to evaluate others' faces based on limited information. This study investigates the cognitive bias in judging the attractiveness of faces partially hidden by sanitary masks. Experiment 1 revealed that men rated women's faces as more attractive when wearing masks, specifically in the context of rating women as romantic partners; however, this mask bias was absent when men rated women as friends. On the other hand, women did not show the mask bias irrespective of the assumed social relationship. Experiment 2 demonstrated that the mask bias among elderly men was less affected by the assumed social relationship (or the possibility of reproduction), compared to young men, though they showed the bias itself. These results suggest that the cognitive strategies related to reproduction underlie the attractiveness judgment of the partial faces.

Acute effects of exercise on gaze fixation and affective response inhibition in children with autism spectrum disorder: A randomized cross-over study.

Children with autism spectrum disorder (ASD) show impairments in response inhibition, especially in socio-emotional contexts. A single aerobic exercise session has the potential to temporarily reduce such impairments as findings from neurotypical children support acute benefits of this exercise type for inhibitory control and emotion recognition. In children with ASD, we therefore aimed to investigate the effects of an aerobic exercise bout on response inhibition in an emotional Go/NoGo task and gaze fixation as possible mechanism underlying changes in performance. Using a cross-over design, 29 patients completed a 20-min aerobic exercise bout at moderate intensity on a cycling ergometer and a control condition in a randomized order. An emotional Go/NoGo task was administered before and after both experimental conditions. Eye-tracking was performed during the cognitive task to assess the duration of gaze fixation of eyes and mouth parts of faces expressing happy or sad emotions. The results support no beneficial effect of exercise on performance on the emotional Go/NoGo task. Instead, patients showed a greater decrease in accuracy on Go trials displaying happy faces in the exercise compared to the control condition. This change was associated with a more pronounced decrease in the fixation duration of the eyes for faces expressing either happy or sad emotions. In conclusion, while a single session of moderately intense aerobic exercise does not affect response inhibition, it temporarily aggravates ASD-specific deficits in the processing of and response to facial emotions.

Reticular Chemistry and In Situ "One-Pot" Strategy: A Dream Combination to Construct Metal-Organic Frameworks.

The establishment of reticular chemistry has significantly facilitated the development of porous materials, especially for metal-organic frameworks (MOFs). On the other hand, as an alternative approach, in situ "one-pot" strategy has been explored as a promising approach to constructing MOFs, in which the synthesis of organic linkers and the sequential construction of MOFs are integrated into one solvothermal condition. This strategy can efficiently avoid the limitations faced in the traditional construction method, such as time-consuming organic synthesis and multiple separation and purification. Herein, inspired by the reaction of aldehydes and o-phenylenediamine and deep structural analysis of UiO-68, a series of tetra-, hexa-, and octa-topic carboxylic acids are synthesized using 2',3'-diamino-[1,1':4',1'"-terphenyl]-4,4'"-dicarboxylic acid and di-, tri-, and tetra-topic aldehydes as precursor. Then nine multicarboxylate-based zirconium MOFs (Zr-MOFs) are successfully constructed via the combination of reticular chemistry and in situ "one-pot" strategy. The resultant Zr-MOFs can be regarded as the partial face decoration of UiO-68. More importantly, the emission properties of resultant Zr-MOFs can be well controlled using aldehydes with tunable electronic structures. This work provides a new path to rational design and construction of porous materials with specific structures guided by reticular chemistry and conducted using in situ "one-pot" strategy.