Tadqiqot nutqni avtomatik tanib olish (NATO)ning nutqni avtomatik tanib olishning neyron tarmoqlariga bag‘ishlangan. O‘tkazilgan tajriba ma'lum test to‘plami tanlanmalaridan foydalanib, nutqni tanib olishda bir nechta ilovalar taqqoslangan. Ma'lumotlar to‘plami har bir tizim tomonidan Python dasturlash tili ilovalari yordamida tahlil qilindi, chiqish ma'lumotlari normallashtiriladi va WER standartiga muvofiq oldindan transkripsiya qilingan etalon ma'lumotlari bilan taqqoslanadi. Sinov natijalari tahlili o‘tkazilgan, nutqni avtomatik tanib olish tizimining samaradorligi uning elementlarini optimallashtirish va kerakli ma'lumotlar to‘plamidan foydalangan holda o‘qitishga bog‘liqligi to‘g‘risida xulosalar chiqarilgan.
Timely detection and recognition of road signs is very important for motorists and pedestrians. The necessary information is delivered to the driver through the program of direct recognition of road signs. There are different types of road signs, some signs may have multi-directional points to indicate the direction of the location, and some signs are about safety rules and restrictions. can do in the message. Identifying and recognizing road signs is an important practice because it helps the new driver to find the way. We are able to solve these problems through the road sign detection program we created in python. In this program, in the following model, each pixel in the binary matrix is 8 bits, so instead of converting their values to 0 and 1, they can be converted to values between 0 and 255. In this way, 128x128 pixel images are converted into binary matrices, and these matrices are prepared for high-level data acquisition models, for example, the CNN model.
Insonni xatti-harakatlariga qarab tanib olish kompyuterni ko’rish va sun’iy intellektning qiziqarli va qiyin sohasidir. So’nggi yillarda odamlarning harakatlari va imo-ishoralarini tahlil qilish orqali aniqlay oladigan modellarni ishlab chiqishda sezilarli yutuqlarga erishildi. Ushbu sohadagi ikkita mashhur model COCO (Kontekstdagi umumiy ob’ektlar) va MPII (Maks Plank Informatika Instituti) modellaridir. Ushbu maqolada biz insonni o’z harakatlaridan tanib olish modellarini o’rganamiz va keyin COCO va MPII qiyosiy tahlilini o’rganamiz, ularning farqlari va kuchli tomonlarini ta’kidlaymiz.
Tanib olish modullarini dasturiy jihatdan amalga oshirish hamda nutqni tanib olish moduli ishini sifatini baholash ishlarini olib borish. Nutq signaliga dastlabki ishlov berish va ularni neyron tarmoqlarida oʻqitishga tayyorlash jarayonini avtomatlashtiruvchi dasturiy modul ishlab chiqildi. Ushbu dasturiy modul yordamida katta xajmdagi nutq maʼlumotlarini tarmoqga kirish standartiga moslash imkoniyatini beradi.
In this paper, local directional pattern (LDP) based methods for frontal face recognition are discussed (summarized). LDP based face feature extraction and comparison methods and their performance results are given. Although, in the paper method for normalizing illuminations of face images is given. Were performed the results of experimental research of the developed algorithms
Respiratory allergosis includes diseases in the pathogenesis of which allergies play a decisive role, and in the clinical picture, symptoms of respiratory damage come to the fore. Allergic lesions of the respiratory system are quite common in children, especially early and preschool age. It is believed that allergic diseases of the nose, pharynx and ear account for 50-60% of all cases of diseases of these organs in children [1, 5, 7]. In recent years, the doctrine of the reactivity of the organism has been successfully developed [2, 4]. Allergic reactivity is understood as the development of a specific hypersensitivity of the body to the effects of genetically alien substances - allergens. Determination of the allergic reactivity of the organism of patients is of great practical importance [3, 6]. In this regard, there is a need for detailed studies related to clarifying the features of the allergic reactivity of the body in various diseases.
Hospital or nosocomial infections are the most common complication in patients in intensive care units and the leading cause of death in both surgical and somatic patients. Despite the presence of a large number of antibacterial drugs in the doctor's arsenal, the results of HI treatment remain unsatisfactory. In recent years, there has been a steady trend towards an increase in the resistance of hospital strains of microorganisms to the most commonly used antibacterial drugs in the clinic. Approximately 90% of all nosocomial infections are caused by bacteria, a distinctive feature of which is resistance to many groups of antibacterial drugs (multiresistance). This is what causes difficulties in the treatment of nosocomial infections, predetermining the low efficiency and high cost of treatment. Resistant strains form under the influence of widely and inappropriately used antibiotics at both prehospital and hospital levels. They can enter the hospital from the body of carrier patients. The transfer of bacteria from patient to patient involves the staff of medical institutions in the process of caring for patients, performing diagnostic procedures, etc. The problem of nosocomial infection, including through respiratory equipment, is very acute due to the increased development of ventilator-associated pneumonia. Along with the impossibility of ensuring the sterilization of anesthesia and respiratory equipment after each patient, there is a serious problem of effective antibacterial therapy of NPV in hospitals.
В статье приведен комплексный анализ сущности слияний и поглощений (M&A) как инструмента укрупнения бизнеса и привлечения инвесторов. Выработаны предложения и решения по совершенствованию действующего законодательства, регулирующего данную сферу.
Sign language recognition has gained significant attention due to its potential to bridge communication gaps between the deaf and hearing communities. This article presents a comprehensive review of machine learning methods employed for the recognition of Uzbek Sign Language (UzSL). The unique visual and spatial nature of sign languages poses challenges that necessitate specialized techniques for accurate recognition. This review surveys various approaches, ranging from traditional techniques to modern deep learning methods, used to recognize UzSL gestures. The article begins by introducing the significance of UzSL recognition and its impact on facilitating effective communication for the Uzbek deaf community. It outlines the complexities involved in sign language recognition, including variations in hand shapes, movements, and facial expressions. The challenges of limited training data, real-time recognition, and capturing dynamic features are discussed in depth. A survey of traditional machine learning methods such as Hidden Markov Models (HMMs), Support Vector Machines (SVMs), and k-Nearest Neighbors (k-NN) is presented, along with their applications and limitations in UzSL recognition. The evolution of these methods into more sophisticated approaches like Dynamic Time Warping (DTW) and Conditional Random Fields (CRFs) is also explored.
In the databases of evidence-based medicine, studies that study the effectiveness of taking vitamin-mineral complexes indicate a high risk of age-related macular degeneration (AMD) as a prevention (primary prevention). The purpose of the study was to evaluate the effectiveness of the use of lutein-zeaxanthin vitamin-mineral complex containing preparations in individuals with a high risk of AMD for the prevention of the disease. The material of the study was 98 individuals (196 eyes) from the 1st (main) group with the highest risk of developing AMD, who agreed to participate in the prevention of AMD and 90 individuals (180 eyes) from the 2nd (control) group, for various reasons refused to take the drug, but agreed to participate in condition monitoring. The follow-up period was 3 years. The results of the observation showed that in persons of the 1st group, there was a stability in the indices of visual acuity and field of vision, ophthalmoscopic and tomographic picture of the macular zone during the entire period of observation. Whereas in persons of the 2nd group by the 3rd year of observation, visual acuity worsened by 2.5 times, the total boundaries of the peripheral visual field narrowed by 47.10, relative and absolute scotomas appeared (p<0.05). The appearance of drusen was observed and in 8 eyes (4.44%) a diagnosis of age-related macular degeneration of the retina, early stage, was made. Conclusions. The proposed scheme of drug prevention of persons with the highest risk of developing AMD (Group 1) showed a significantly positive effect on the functional state of the retina, leads to a stable preservation of visual functions during 3 years of observation and prevents the occurrence of AMD in 100% of individuals.
The monograph presents modern aspects of syndialysis arterial hypotension: solved and unresolved problems of predicting, preventing and diagnosing this complication during dialysis. The characteristics of clinical and pathogenetic features, diagnosis, treatment, prognosis and prevention of syndialysis arterial hypotension are given. The data of the dynamics of clinical, functional and laboratory studies of syndialytic arterial hypotension in patients on dialysis are presented. An algorithm for verifying the status of hydration and preventing syndialysis hypotension is presented. The monograph is intended for nephrologists, doctors of related specialties, masters and students of medical institutes.