Search

Visual speech recognition or automated lip-reading systems actively apply to speech-to-text translation. Video data proves to be useful in multimodal speech recognition systems, particularly when using acoustic data is difficult or not available at all. The main purpose of this study is to improve driver command recognition by analyzing visual information to reduce touch interaction with various vehicle systems (multimedia and navigation systems, phone calls, etc.) while driving. We propose a method of automated lip-reading the driver’s speech while driving based on a deep neural network of 3DResNet18 architecture. Using neural network architecture with bi-directional LSTM model and attention mechanism allows achieving higher recognition accuracy with a slight decrease in performance. Two different variants of neural network architectures for visual speech recognition are proposed and investigated. When using the first neural network architecture, the result of voice recognition of the driver was 77.68 %, which was lower by 5.78 % than when using the second one the accuracy of which was 83.46 %. Performance of the system which is determined by a real-time indicator RTF in the case of the first neural network architecture is equal to 0.076, and the second — RTF is 0.183 which is more than two times higher. The proposed method was tested on the data of multimodal corpus RUSAVIC recorded in the car. Results of the study can be used in systems of audio-visual speech recognition which is recommended in high noise conditions, for example, when driving a vehicle. In addition, the analysis performed allows us to choose the optimal neural network model of visual speech recognition for subsequent incorporation into the assistive system based on a mobile device.

Automatic audio-visual speech recognition systems (AVSRs) have recently achieved tremendous success, especially in limited vocabulary tasks by far surpassing human abilities to recognize speech, especially in acoustically noisy conditions. Automatic speech recognition systems based on processing of audio and video information are being actively researched and developed all over the world. However, scientific studies aimed at analyzing the influence of the speaker's emotional state (anger, disgust, fear, happy, neutral, and sad), and, most importantly, intensity level of emotion (low - LO, medium - MD, high - HI) on automatic lip-reading have not been conducted. In this regard, the relevance of this research topic cannot be overestimated and requires detailed study. In this paper, we present a novel approach for emotional speech lip-reading, that includes evaluation of a speaker’s emotion and its intensity level. The proposed approach uses visual speech data to detect a person’s emotion type and its intensity level and based on this information assigns it to one of the trained emotional lip-reading models. This essentially resolves the multi-emotional lip-reading issue associated with most real-life scenarios. The proposed approach improves the state-of-the-art results due to the consideration of the intensity of the pronounced audio-visual speech up to 8.2% in terms of the accuracy. Current research is the first step in the creation of emotion-robust speech recognition systems and leaves open a wide field for further research.

Personality and affective computing techniques play a significant role for better understanding of human’s behavior and intentions. Such techniques can be applied in practice in recommendation systems, healthcare, education, and job applicant screening. In this paper, we propose a novel multimodal approach to personality traits assessment that leverages affective features of human’s voice and face, as well as recent advances in the deep learning. We present a new mid-level modality fusion strategy that is based on a cross-modal attention mechanism with summarizing functionals. In contrast to other state-of-the-art approaches, we not only analyze a visual scene, but specifically process human’s upper body (selfie) and a scene background. Our experiments show that the Extroversion personality trait is better estimated by fusing visual scene, face, and audio (voice) modalities, while the Conscientiousness and Agreeableness traits are better assessed by fusing face, selfie, and audio modalities. Furthermore, our results show that utilizing the selfie modality outperforms the visual scene modality by more than 1% in terms of the Concordance Correlation Coefficient. Additionally, our approach based on processing three modalities (selfie, face, and audio) is on-par with other known state-of-the-art approaches that employ at least four modalities on the test set of the ChaLearn First Impressions V2 corpus.

В статье представлен аналитический обзор исследований в области аффективных вычислений. Это направление является составляющей искусственного интеллекта, и изучает методы, алгоритмы и системы для анализа аффективных состояний человека при его взаимодействии с другими людьми, компьютерными системами или роботами. В области интеллектуального анализа данных под аффектом подразумевается проявление психологических реакций на возбуждаемое событие, которое может протекать как в краткосрочном, так и в долгосрочном периоде, а также иметь различную интенсивность переживаний. Аффекты в рассматриваемой области разделены на 4 вида: аффективные эмоции, базовые эмоции, настроение и аффективные расстройства. Проявление аффективных состояний отражается в вербальных данных и невербальных характеристиках поведения: акустических и лингвистических характеристиках речи, мимике, жестах и позах человека. В обзоре приводится сравнительный анализ существующего информационного обеспечения для автоматического распознавания аффективных состояний человека на примере эмоций, сентимента, агрессии и депрессии. Немногочисленные русскоязычные аффективные базы данных пока существенно уступают по объему и качеству электронным ресурсам на других мировых языках, что обуславливает необходимость рассмотрения широкого спектра дополнительных подходов, методов и алгоритмов, применяемых в условиях ограниченного объема обучающих и тестовых данных, и ставит задачу разработки новых подходов к аугментации данных, переносу обучения моделей и адаптации иноязычных ресурсов. В статье приводится описание методов анализа одномодальной визуальной, акустической и лингвистической информации, а также многомодальных подходов к распознаванию аффективных состояний. Многомодальный подход к автоматическому анализу аффективных состояний позволяет повысить точность распознавания рассматриваемых явлений относительно одномодальных решений. В обзоре отмечена тенденция современных исследований, заключающаяся в том, что нейросетевые методы постепенно вытесняют классические детерминированные методы благодаря лучшему качеству распознавания состояний и оперативной обработке большого объема данных. В статье рассматриваются методы анализа аффективных состояний. Преимуществом использования многозадачных иерархических подходов является возможность извлекать новые типы знаний, в том числе о влиянии, корреляции и взаимодействии нескольких аффективных состояний друг на друга, что потенциально влечет к улучшению качества распознавания. Приводятся потенциальные требования к разрабатываемым системам анализа аффективных состояний и основные направления дальнейших исследований.

Introduction. Prevention and treatment of periodontal diseases remains one of the urgent problems of modern medicine. According to WHO, the prevalence of dental diseases among adults is 90–95 %, among adolescents – 80–83 %. The health of the body as a whole largely depends on the state of the oral cavity. It is important for the effective treatment of diseases to have a properly selected dosage form, which provides both convenience of use and purposeful use of the action of the pharmacologically active drug contained in it. Medicinal pencils as a dosage form are very limited on the pharmaceutical market, so their development is relevant. As a local anesthetic component, anilocaine is included in the composition of pencils at a concentration of 5 %, antimicrobial action is provided by the introduction of metronidazole in a concentration of 1 % in order to prevent the occurrence of infection.Aim. Substantiation of the composition and technology of the medicinal pencil for the treatment and prevention of infectious and inflammatory diseases of the oral cavity.Materials and methods. For the study, we used medicinal pencils with an active pharmaceutical substance – anilocaine (FS 42-2846-97, LLC "Perm Chemical Company", Perm, shelf life of 5 years), metronidazole (FS 42-0257-07, JSC "Medisorb", Perm, shelf life of 3 years); basic components – PEO-400 (FS 42-1242-96, LLC "Neftegazkhimkomplekt", Moscow, shelf life of 3 years), PEO-1500 (FS 42-1885-96, LLC "Neftegazkhimkomplekt", Moscow, shelf life of 3 years), PEO-4000 (FS 42-3337-96, LLC "Neftegazkhimkomplekt", Moscow, shelf life 3 years), second ingredient – purified water (FS.2.2.0020.18), the UVOI-"MF" 1812S6 obtained at the reverse osmosis plant (JSC "NPK Median-Filter", Russia). In the course of the research, technological, structural-mechanical and biopharmaceutical methods were used.Results and discussion. Based on technological, structural-mechanical and biopharmaceutical studies, the composition of medicinal pencils with anesthetic, anti-inflammatory and antimicrobial effects for the treatment and prevention of infectious and inflammatory diseases of the oral cavity is proposed and justified for the first time. Biopharmaceutical availability of medicinal pencils is determined by the conductometric method.Conclusion. The conducted complex of studies will expand the range of domestic medicines for the treatment and prevention of infectious and inflammatory diseases of the oral cavity.

Monitoring and evaluation of the safety level of individuals is one of the most important problems of the modern world, which was forced to change due to the emergence of the COVID-19 virus. To increase the safety level of individuals, new information technologies are needed that can stop the spread of infection by minimizing the threat of outbreaks and monitor compliance with recommended measures. These technologies, in particular, include intelligent tracking systems of the presence of protective face masks. For these systems, this article proposes a new method for generating training data that combines data augmentation techniques, such as Mixup and Insert. The proposed method is tested on two datasets, namely, the MAsked FAce dataset and the Real-World Masked Face Recognition Dataset. For these datasets, values of the unweighted average recalls of 98.51% and 98.50% are obtained. In addition, the effectiveness of the proposed method is tested on images with face mask imitation on people's faces, and an automated technique is proposed for reducing type I and II errors. Using the proposed automated technique, it is possible to reduce the number of type II errors from 174 to 32 for the Real-World Masked Face Recognition Dataset, and from 40 to 14 for images with painted protective face masks.

The problem of masked face recognition is investigated. It is shown that real masks of various shapes, textures and colors have become a problem for state-of-the-art face recognition systems. A reason for this is the lack of the necessary real training datasets. Creation of new data based on simple methods of forming masks on face images could solve this problem. An original method is proposed including the generation of various types, shapes, and colors of masks directly on the original texture of face images. The formation of the masks on the faces of individuals, on faces in group photos, and in scenes with streams of people was taken into account. Based on 100 original face images from the CUHK Face Sketch Database, a test database was created that includes more than 20,000 masked faces images which available for use. Experiments were carried out to recognize faces from the test database within the implemented four systems, among which three are state-of-the-art systems based on “deep learning” and one is deterministic system based on the cosine-transform. The performance of these systems was evaluated, the obtained results of masked face recognition were interpreted, and the masks that were a problem for selected four systems were noted. The proposed mask generation method can be used to create corpora and test databases of images with masks. The obtained results will be useful to researchers and specialists in the field of image processing and analysis.

In this article, we present a novel approach for emotional speech lip-reading (EMOLIPS). This two-level approach to emotional speech to text recognition based on visual data processing is motivated by human perception and the recent developments in multimodal deep learning. The proposed approach uses visual speech data to determine the type of speech emotion. The speech data are then processed using one of the emotional lip-reading models trained from scratch. This essentially resolves the multi-emotional lip-reading issue associated with most real-life scenarios. We implemented these models as a combination of EMO-3DCNN-GRU architecture for emotion recognition and 3DCNN-BiLSTM architecture for automatic lip-reading. We evaluated the models on the CREMA-D and RAVDESS emotional speech corpora. In addition, this article provides a detailed review of recent advances in automated lip-reading and emotion recognition that have been developed over the last 5 years (2018–2023). In comparison to existing research, we mainly focus on the valuable progress brought with the introduction of deep learning to the field and skip the description of traditional approaches. The EMOLIPS approach significantly improves the state-of-the-art accuracy for phrase recognition due to considering emotional features of the pronounced audio-visual speech up to 91.9% and 90.9% for RAVDESS and CREMA-D, respectively. Moreover, we present an extensive experimental investigation that demonstrates how different emotions (happiness, anger, disgust, fear, sadness, and neutral), valence (positive, neutral, and negative) and binary (emotional and neutral) affect automatic lip-reading.

Introduction. Soft gelatin capsules are a promising dosage form comprising essential oils as active agents. Joint research of the staff of the Perm State Pharmaceutical Academy, the Research Institute of Nutrition of the Ministry of Energy and Industry of the Republic of Tajikistan have proposed the composition of gelatin mass for encapsulation by the rotary matrix method. The mechanical and physical-technological parameters required to preserve the strength and elasticity of the capsule shell during the production process and storage are determined.Aim. Study of the rheological properties of gelatin masses based on gelatin of different grades, as well as the migration of essential oils through capsule shells.Materials and methods. Pharmaceutical active substances "Lipovitol" and "Limoneol" obtained in the Republic of Tajikistan were used as active substances introduced into the composition of soft gelatin capsules. Sunflower oil was used as the solvent. Gelatin samples were used to obtain gelatin masses: 1 – Foodchem (China), 2 – Brodnickie Zaklady Zelatyny Sp. zo.o. (Poland), 3 - Italgelatine s.p.a. (Italy), 4 – Ewald-Gelatine GmbH (Germany), 5 – Weishardt International (France); glycerol; sunflower oil. Gelatin mass for manufacturing soft capsules was prepared in a closed reactor. Capsules were prepared on an automatic encapsulation line RJWJ – 115 Soft Gelatin Encapsulator Machine (China). The structural and mechanical properties of soft gelatin masses were determined on a rotary viscometer RV type "Reotest 2" (Germany). The dynamics of the process of migration of essential oils and its components were studied by changing their amount in a capsule by chromato-mass spectrometry method on a chromatograph Varian CP 3800 with a quadrupole mass spectrometer 4000 MS as a detector (USA).Results and discussion. When studying the rheological properties of model compositions, it was found that for all samples of gelatin masses there is a decrease in values of effective viscosity when the shear rate increases, which characterizes the tested samples as a structured dispersion system. Additional studies have shown that the gelatin masses have thixotropic properties. Samples of gelatin masses 3–5 had narrower hysteresis loops, while sample 5 the narrowest, restoration of the structure took place quite quickly. Capsules obtained from gelatin mass 3 and 4 samples had a strong seam and were well cut out of the tape. From the mass of sample 5, high strength ribbons were obtained, a high temperature was required to seal the capsules, in some capsules the seam was not glued on one side, as a result, the capsules were rigid and brittle. As a result of the study, the rheological optimum of the gelatin mass suitable for preparing capsules by a rotary matrix method was determined, which has boundaries in the ranges of shear rates of 0.556–243 s-1 and viscosity ranges of 11.46-5028.76 Pa ⋅ s and shear stress of 2788–2808 Pa developing at these rates. When studying the migration of active substances through the capsule shell, it was found that over three years of storage of capsules in a closed polymer can, the content of essential oil in Lipovitol capsules decreased by 4.88 %, in Limoneol capsules by 5 %, which indicates partial migration of oil through the gelatin shell. The content remained within the permissible deviations (±10 %). The content of essential essential oil components also remained within acceptable deviations throughout the shelf life.Conclusion. The optimal composition of the shell for producing soft capsules by a rotary matrix method is justified. It was found that the rheological optimum of gelatin mass is characterized by viscosity ranges of 11.46-5028.76 Pa ⋅ s and shear stress of 2788–2808 Pa. According to the results of the study of the migration of essential oils through the shell, has been established the shelf life of soft gelatin capsules in glass jars made of dark glass and a temperature of 15 to 25 °C – 3 years.

Начиная с 2019 года все страны мира столкнулись со стремительным распространением пандемии, вызванной коронавирусной инфекцией COVID-19, борьба с которой продолжается мировым сообществом и по настоящее время. Несмотря на очевидную эффективность средств индивидуальной защиты органов дыхания от заражения коронавирусной инфекцией, многие люди пренебрегают использованием защитных масок для лица в общественных местах. Поэтому для контроля и своевременного выявления нарушителей общественных правил здравоохранения необходимо применять современные информационные технологии, которые будут детектировать защитные маски на лицах людей по видео- и аудиоинформации. В статье приведен аналитический обзор существующих и разрабатываемых интеллектуальных информационных технологий бимодального анализа голосовых и лицевых характеристик человека в маске. Существует много исследований на тему обнаружения масок по видеоизображениям, также в открытом доступе можно найти значительное количество корпусов, содержащих изображения лиц как без масок, так и в масках, полученных различными способами. Исследований и разработок, направленных на детектирование средств индивидуальной защиты органов дыхания по акустическим характеристикам речи человека пока достаточно мало, так как это направление начало развиваться только в период пандемии, вызванной коронавирусной инфекцией COVID-19. Существующие системы позволяют предотвратить распространение коронавирусной инфекции с помощью распознавания наличия/отсутствия масок на лице, также данные системы помогают в дистанционном диагностировании COVID-19 с помощью обнаружения первых симптомов вирусной инфекции по акустическим характеристикам. Однако, на сегодняшний день существует ряд нерешенных проблем в области автоматического диагностирования симптомов COVID-19 и наличия/отсутствия масок на лицах людей. В первую очередь это низкая точность обнаружения масок и коронавирусной инфекции, что не позволяет осуществлять автоматическую диагностику без присутствия экспертов (медицинского персонала). Многие системы не способны работать в режиме реального времени, из-за чего невозможно производить контроль и мониторинг ношения защитных масок в общественных местах. Также большинство существующих систем невозможно встроить в смартфон, чтобы пользователи могли в любом месте произвести диагностирование наличия коронавирусной инфекции. Еще одной основной проблемой является сбор данных пациентов, зараженных COVID-19, так как многие люди не согласны распространять конфиденциальную информацию.

Recognition of emotions and sentiment (affective states) from human audio–visual information is widely used in healthcare, education, entertainment, and other fields; therefore, it has become a highly active research area. The large variety of corpora with heterogeneous data available for the development of single-corpus approaches for recognition of affective states may lead to approaches trained on one corpus being less effective on another. In this article, we propose a multi-corpus learned audio–visual approach for emotion and sentiment recognition. It is based on the extraction of mid-level features at the segment level using two multi-corpus temporal models (a pretrained transformer with GRU layers for the audio modality and pre-trained 3D CNN with BiLSTM-Former for the video modality) and on predicting affective states using two single-corpus cross-modal gated self-attention fusion (CMGSAF) models. The proposed approach was tested on the RAMAS and CMU-MOSEI corpora. To date, our approach has outperformed state-of-the-art audio–visual approaches for emotion recognition by 18.2% (78.1% vs. 59.9%) for the CMU-MOSEI corpus in terms of the Weighted Accuracy and by 0.7% (82.8% vs. 82.1%) for the RAMAS corpus in terms of the Unweighted Average Recall.

This paper proposes a new hybrid method for automatic detection and recognition of the presence/absence of a protective mask on human's face. It combines visual features extracted using Convolutional Neural Network (CNN) with image histograms that convey information about pixel intensity. Several pre-trained models for building feature extraction systems using a CNN and several types of image histograms are considered in this paper. We test our approach on the Medical Mask Dataset and perform cross-corpus analysis on two other databases named Masked Faces (MAFA) and Real-World Masked Face Dataset (RMFD). We demonstrate that the proposed hybrid method increases the Unweighted Average Recalls (UARs) of recognition of the presence/absence of a protective mask on human's face in comparison with traditional CNNs on the MAFA and RMFD databases by 0.96% and 1.32%, respectively. The proposed method can be generalized and used for other tasks of biometry, computer vision, machine learning and automatic face recognition.

Introduction. The study is of relevance as the influence of environmental factors is growing in all socio-economic and socio-psychological processes of modern society. Based on the study conducted, the article analyzes the processes of interaction between environmental factors and the mentality of a person and society. Materials and Methods. Works of Russian and foreign researchers on the interdependence of mentality and the environmental situation, as well as the authors' own long-term studies of this relationship were used as materials for the study. Particular attention was paid to such factors as natural resources, hypertrophied consumption, pollution and degradation of the natural environment. When examining the issue of anthropogenic destruction of the biosphere and the relationship of these processes with mentality, the ecological characteristics of the current stage of world development, specific regional problems of Russia and other countries and regions were considered. The main research method employed was the systemic socio-eco-economic analysis of situations and the systematic account of the key characteristics of regional specificity. Results. Dependence of mentality on regional environmental factors has been revealed. The processes of formation of the environmental mentality in various social groups, the regularities of the influence of the mentality of the population, business, government agencies on the emergence and development of environmental problems have been analyzed. The case study of Russia’s regions with the most acute environmental problems has shown that it is the traditional mentality that is largely the cause of the emergence and escalation of conflict situations. The principles of consistent environmentalization of the mentality of the population, business, federal and regional government agencies have been stated. Discussion and Conclusion. The idea that it is the mentality of the human consumer and that of the mass consumption society that is steadily leading to an environmental disaster is gaining the understanding very slowly. There is a need for increasing efforts to systematically environmentalize the mentality when perceiving world, regional, and local problems. The article will be useful to a wide range of specialists in the analysis and synthesis of socio-eco-economic systems. The results of the study may be applied for prevention and de-escalation of regional and interregional conflicts associated with environment and nature conservation.

Mother’s milk is the most suitable product for feeding a newborn baby, its uniqueness is obvious. However, there are situations when breastfeeding is not possible at all or its volume is limited, so under certain circumstances mothers are forced to partially supplement or completely feed the child with an artificial formula. Although the composition of the ingredients in most modern formula feeds is similar, there are nevertheless differences in their composition and in the processing of the ingredients in the formula. In addition, the child may have individual taste preferences as well as unexpected negative reactions to any adapted milk formula. This review presents the achievements of recent years in creating adapted goat’s milk-based formulas for feeding newborns and first-year infants. Goat’s milk-based formulas have a number of advantages over cow’s milk-based formulas. One of the representatives of modern artificial formulas are those containing goat’s milk, which is characterized by low α-s1-casein genotype and high β-lactoglobulin content, which provides easier protein digestibility, contributes to more comfortable digestion and better absorption. Knowledge of the biological and nutritional value of goat milk, taking into account the peculiarities of the gastrointestinal tract of newborns and their needs, made it possible to develop and introduce into the goat’s milk-based formula the DigestX lipid complex, which is a complex of vegetable oils with a high content of palmitic acid in the sn-2-position in the triglyceride molecule, which is similar to breast milk. Goat’s milk, compared to cow’s milk, contains 4-6 times more oligosaccharides, which retain their beneficial properties in the production of artificial goat’s milk formula.Recent scientific literature data is presented, including that from the standpoint of evidence-based medicine, proving the effectiveness and safety of artificial goat’s milk-based formulas. The necessity of consulting mothers, psychological support and practical assistance not only in breastfeeding, but also in artificial feeding is substantiated.

Hydrological and chemical surveys were conducted in the Uglovoy Bight in October, 2019, February, May and June, 2020 (in total 120 stations) and chemical analyses of water from 13 small rivers running into the bight were done on October 21-22, 2020. Extremely high concentration of nutrients was detected in the Peschanka, Saperka and Gryaznukha Rivers that was obviously caused by waste waters discharge. These rivers were the main source of the bight eutrophication. Within the bight, the highest anomalies of chemical parameters, as low oxygen content, low pH, high concentrations of nutrients (N, P, Si), high turbidity, and high CO2 partial pressure were observed close to these rivers mouths, in particular under the ice in winter, when wind mixing was absent. The hypoxia disappeared in the warm period of year because of wind mixing. High concentrations of total nitrogen (10.0-40.0 μmol/L), total phosphorus (1.5-2.0 gmol/L), dissolved organic carbon (3-5 mgC/L), and chlorophyll a (0.5-2.0 μg/L) in all seasons were the results of active production-destruction processes, obviously with prevalence of organic matter destruction, since the water in the bight was undersaturated with oxygen and supersaturated with carbon dioxide — the bight accumulates and mineralizes organic matter from terrestrial and riverine discharge. Underwater photographs did not detect Zostera meadows at the bottom, which were observed in the northwestern Uglovoy bight in the past. Comparison of historical data on episodic studies in the bight with results of the surveys indicates degradation of its ecosystem, with such signs as disappearance of seagrass, hypoxia in winter, and CO2 flux into the atmosphere. Reduce in water exchange between the bight and the Amur Bay caused by construction of the underwater pipeline in 1982 and the bridge in 2012 is suggested as a reason of the degradation.

Subject of Research. The imbalance of classes in datasets has a negative impact on machine classification systems used in applications of artificial intelligence, such as: medical diagnostics, fraud detection and risk management. This problem in facial expression datasets also degrades the performance of classification algorithms. Method. The paper discusses the main approaches for the class imbalance reduction: resampling methods and setting the weight of classes depending on the number of samples observed for an each class. A histogram of oriented gradients is used for the face area localization in the frame stream, then an active shape model is applied, which detects the coordinates of 68 key facial landmarks. Using the coordinates of key landmarks, informative features are extracted that characterize the dynamics of facial expressions. Main Results. The results of the study have shown that the proposed approach to the extraction of visual features exceeds the accuracy of human emotion recognition by facial expressions. The considered methods of the class imbalance reduction in the set of facial expressions have provided the improvement of machine classifier performance and showed that the existing class imbalance in a training set has a significant effect on the accuracy. Practical Relevance. The proposed approach to the extraction of visual features can be used in automatic systems for human emotion recognition by facial expressions, and result analysis of applying methods that reduce class imbalance can be useful for researchers in the field of machine learning.

Recognition of human emotions by facial expressions is an important research problem that covers many areas and disciplines, such as computer vision, artificial intelligence, medicine, psychology and security. This paper provides an analytical overview of video facial expression databases and approaches to recognition emotions by facial expressions, which include three main stages of image analysis, such as pre-processing, feature extraction and classification. The paper presents both traditional approaches to recognition of human emotions by visual facial features, and approaches based on deep learning using deep neural networks. We give the current results of some existing algorithms. In the review of scientific and technical literature we empathized mainly the sources containing theoretical and research information of the methods under consideration, as well as comparison of traditional methods and methods based on deep neural networks, which were supported by experimental studies. Analysis of scientific and technical literature describing methods and algorithms for study and recognition of facial expressions, as well as the results of world scientific research, have shown that traditional methods for classification of facial expressions are second in speed and accuracy to artificial neural networks. The main contribution of this review is providing a common understanding of modern approaches to recognition of facial expressions, which will enable new researchers to understand the main components and trends in the field of recognition of facial expressions. Moreover, comparison of world scientific findings has shown that a combination of traditional approaches and approaches based on deep neural networks achieves better classification accuracy, but artificial neural networks are the best classification methods. The paper may be useful to specialists and researchers in the field of computer vision.

Breastfeeding is one of the main components of the health and optimal growth of a newborn child. The article presents recommendations on breastfeeding support in case of lactostasis and mastitis, as well as the role of neonatologist and pediatrician in prevention and preservation of breastfeeding when mother faces problems, lactation support in case of lactational mastitis development in mother. Prolonged lactostasis may be the cause of mastitis - breast inflammation, which is characterized by painfulness, swelling and redness of the breast. If the symptoms persist for 12-24 hours, in conditions of impaired milk evacuation the growth of microorganisms occurs, which leads to infectious lactation mastitis, which may be complicated by an abscess. The main risk factors of mastitis development, various methods of diagnostics, lactostasis and lactation mastitis prophylaxis are considered, and the special role of breastfeeding advice to the mother is emphasized. Data on modern and prospective method of lactostasis and mastitis prophylaxis - application of the product Lactanza based on Lactobacillus fermentum CECT5716 are presented. The use of probiotic bacteria was proposed as a method of modulating the immune system of a developing child to reduce the risk of immune aberrations and improve the immune protection of his or her body. According to the published data, there was a 51% reduction of mastitis frequency in the group that received probiotic Lactobacillus fermentum CECT5716. Therefore, it is advisable for women who start breastfeeding with increased risk factors for lactostasis and mastitis to recommend taking Lactobacillus fermentum CECT5716 probiotic immediately after childbirth during the first month of breastfeeding. The drug is also indicated for treatment of lactostasis in breastfeeding women to prevent mastitis from developing and after antibiotic therapy of mastitis to restore breast microbiome and lower the risk of relapse.