Search

In this paper, we introduce a large scale dataset, called HKR, to address challenging detection and recognition problems of handwritten Russian and Kazakh text in the scanned documents. We present a new Russian and Kazakh database (with about 95% of Russian and 5% of Kazakh words/sentences respectively) for offline handwriting recognition. A few pre-processing and segmentation procedures have been developed together with the database. The database is written in Cyrillic and shares the same 33 characters. Besides these characters, the Kazakh alphabet also contains 9 additional specific characters. This dataset is a collection of forms. The sources of all the forms in the datasets were generated by LaTeXwhich subsequently was filled out by persons with their handwriting. The database consists of more than 1500 filled forms. There are approximately 63000 sentences, more than 715699 symbols produced by approximately 200 different writers. It can serve researchers in the field of handwriting recognition tasks by using deep and machine learning. For experiments, we used several popular text recognition methods for word and line recognition like CTC-based and attention-based methods. The results indicate the diversity of HKR. The dataset is available at https://github.com/abdoelsayed2016/HKR_Dataset .

A mixed inverse problem for determining the biochemical oxygen demand of water ( L 0 ) and the rate of biochemical oxygen consumption ( k 0 ), which are important indicators of water quality, has been formulated and numerically solved based on real experimental data. The inverse problem is reduced to the optimization problem consisting in minimization of the deviation of the calculated values from the experimental data, which is solved numerically using the Nelder–Mead method (zero order) and the gradient method (first order). A number of examples of processing both model experimental data and field experimental data provided by hydrological stations monitoring pollutants in the Kazakhstani part of the Ili River basin are presented. A mathematical model that adequately describes the processes in the river system has been constructed.

This article discusses the problem of handwriting recognition in Kazakh and Russian languages. This area is poorly studied since in the literature there are almost no works in this direction. We have tried to describe various approaches and achievements of recent years in the development of handwritten recognition models in relation to Cyrillic graphics. The first model uses deep convolutional neural networks (CNNs) for feature extraction and a fully connected multilayer perceptron neural network (MLP) for word classification. The second model, called SimpleHTR, uses CNN and recurrent neural network (RNN) layers to extract information from images. We also proposed the Bluechet and Puchserver models to compare the results. Due to the lack of available open datasets in Russian and Kazakh languages, we carried out work to collect data that included handwritten names of countries and cities from 42 different Cyrillic words, written more than 500 times in different handwriting. We also used a handwritten database of Kazakh and Russian languages (HKR). This is a new database of Cyrillic words (not only countries and cities) for the Russian and Kazakh languages, created by the authors of this work.

This article presents the Software as a service system that allows creating scenario modeling of pollution transfer for diffuse sources of pollution using the example of the Ili river basin (Republic of Kazakhstan). The development of technologies that determine hydrological state of the lake and water in it are analyzed. The practicability of modeling the discharges and distribution of pollutants is substantiated. The FORTRAN Hydrological Simulation Program software, a computer model that allows us to model the concentration of nitrate compounds (NO3), total ammonium, and biochemical oxygen consumption with one day time resolution for the period from 1980 to 2016 was described. The model was calibrated using the field observations data from 6 hydrological posts, which made it possible to obtain satisfactory water discharge values. To work with the system, a graphical interface was developed which allows the user who is not familiar with the FORTRAN Hydrological Simulation Program software to make calculations. Implemented was an algorithm for automated starting of scenario calculations with post-processing and presentation of results. The web-based approach facilitates multi-user, one-time and fast access to the system from anywhere in the world. The efficiency of results of programming was investigated and the dynamics of changes after using the FORTRAN Hydrological Simulation Program software was established.

This paper demonstrates a qualitative evaluation/comparison of face embeddings extracted from deep learning models, such as VGG-Face, Dlib, and OpenFace, on a face discrimination task. While conducting experiments, each of linear SVM (support vector machine) classifier, Euclidean distance, and Cosine distance algorithms was utilized to compare/analyze face vectors (embeddings) extracted from those 3 deep learning models. This resulted in 9 overall combinations of face recognition techniques to be compared. To implement a fair comparison of these 9 combinations, first, training and test datasets were gathered; these datasets were made up of complete frontal and well-cropped (NxN sized) face images of 33 persons (mostly Asian), with at least 10 different face images for each person. Then, face images in the training dataset were introduced into deep learning models to extract face vectors from them. Next, these face vectors were stored in a local directory as a reference database (to be used with Euclidean and Cosine distance methods) and were used to train SVM classifier. Subsequently, these face vectors were utilized to classify (recognize) face images (vectors) from the test dataset. As experiments proved, the best face recognition technique amongst 9 combinations was Dlib based face recognition model (with SVM classifier combined) as it showed the highest rate to distinguish people from each other.Although, this research work does not bring novelty to the domain, it took an effort to evaluate/compare well-known deep face models performances on Asian faces (mostly) and choose the best one to utilize as a basis for door access control application.

Despite the transition to digital information exchange, many documents, such as invoices, taxes, memos and questionnaires, historical data, and answers to exam questions, still require handwritten inputs. In this regard, there is a need to implement Handwritten Text Recognition (HTR) which is an automatic way to decrypt records using a computer. Handwriting recognition is challenging because of the virtually infinite number of ways a person can write the same message. For this proposal we introduce Kazakh handwritten text recognition research, a comprehensive dataset of Kazakh handwritten texts is necessary. This is particularly true given the lack of a dataset for handwritten Kazakh text. In this paper, we proposed our extensive Kazakh offline Handwritten Text dataset (KOHTD), which has 3000 handwritten exam papers and more than 140335 segmented images and there are approximately 922010 symbols. It can serve researchers in the field of handwriting recognition tasks by using deep and machine learning. We used a variety of popular text recognition methods for word and line recognition in our studies, including CTC-based and attention-based methods. The findings demonstrate KOHTD's diversity. Also, we proposed a Genetic Algorithm (GA) for line and word segmentation based on random enumeration of a parameter. The dataset and GA code are available at https://github.com/abdoelsayed2016/KOHTD.

We present TNCR, a new table dataset with varying image quality collected from free websites. The TNCR dataset can be used for table detection in scanned document images and their classification into 5 different classes. TNCR contains 9428 high-quality labeled images. In this paper, we have implemented state-of-the-art deep learning-based methods for table detection to create several strong baselines. Cascade Mask R-CNN with ResNeXt-101-64x4d Backbone Network achieves the highest performance compared to other methods with a precision of 79.7%, recall of 89.8%, and f1 score of 84.4% on the TNCR dataset. We have made TNCR open source in the hope of encouraging more deep learning approaches to table detection, classification, and structure recognition. The dataset and trained model checkpoints are available at https://github.com/abdoelsayed2016/TNCR_Dataset.

This article considers the task of handwritten text recognition using attention-based encoder&ndash, decoder networks trained in the Kazakh and Russian languages. We have developed a novel deep neural network model based on a fully gated CNN, supported by multiple bidirectional gated recurrent unit (BGRU) and attention mechanisms to manipulate sophisticated features that achieve 0.045 Character Error Rate (CER), 0.192 Word Error Rate (WER), and 0.253 Sequence Error Rate (SER) for the first test dataset and 0.064 CER, 0.24 WER and 0.361 SER for the second test dataset. Our proposed model is the first work to handle handwriting recognition models in Kazakh and Russian languages. Our results confirm the importance of our proposed Attention-Gated-CNN-BGRU approach for training handwriting text recognition and indicate that it can lead to statistically significant improvements (p-value &lt, 0.05) in the sensitivity (recall) over the tests dataset. The proposed method&rsquo, s performance was evaluated using handwritten text databases of three languages: English, Russian, and Kazakh. It demonstrates better results on the Handwritten Kazakh and Russian (HKR) dataset than the other well-known models.

Gravimetry is associated with analysis of the gravitational field. The gravitational field is characterized by its potential. This is described by the Poisson equation, the right side of which includes the density of the environment. There exists direct and inverse problems of gravimetry. Direct gravimetry problems involve the determination of the potential of the gravitational field in a given region. The inverse problems of gravimetry imply the restoration of the structure of a given area from the results of measuring the characteristics of the gravitational field. Such studies are needed to assess on the basis of gravimetric geodynamic events occurring in oil and gas fields. The relevance of such research is necessary, because with prolonged development of the oil and gas fields, negative consequences may occur. This paper discusses some of the features of direct and inverse gravimetry problems. A description of the mathematical model of the processes under consideration is given. Different direct and inverse gravimetry problems are posed. Describes the methods of its solving. Based on the analysis of the results of a computer experiment, appropriate conclusions are made.

This work is devoted to the identification of a mathematical model of bacteria population under the antibiotic influence, based on the solution of the corresponding inverse problems. These problems are solved by the gradient method, genetic algorithm and Nelder–Mead method. Calculations are made using model and real data.

We investigate the continuation problem for the elliptic equation. The continuation problem is formulated in operator form . The singular values of the operator A are presented and analyzed for the continuation problem for the Helmholtz equation. Results of numerical experiments are presented.

We consider the continuation problem from the time-like surface for the 2D Maxwell equation. The problem is formulated in an operator form . We describe and justify gradient methods for minimizing the cost functional for the continuation and coefficient inverse problems. The results of a computational experiment are presented.

A numerical method for solving the Dirichlet problem for the wave equation in the two-dimensional space is proposed. The problem is analyzed for ill-posedness and a regularization algorithm is constructed. The first stage in the regularization process consists in the Fourier series expansion with respect to one of the variables and passing to a finite sequence of Dirichlet problems for the wave equation in the one-dimensional space. Each of the obtained Dirichlet problems for the wave equation in the one-dimensional space is reduced to the inverse problem with respect to a certain direct (well-posed) problem. The degree of ill-posedness of the inverse problem is analyzed based on the character of decreasing of the singular values of the operator A. The numerical solution of the inverse problem is reduced to minimizing the objective functional . The results of numerical calculations are presented.

A nonlinear inverse problem of pharmacokinetics for the dual-compartment model with extravascular route of the drug administration is considered. The problem is significantly ill-posed: its solution is not unique and is unstable. We discuss nonuniqueness of the inverse problem and the degree of ill-posedness. The initial problem is reduced to a linear operator equation by linearization. It is demonstrated that the resolving ability of the inverse problem can be improved by varying of the location of measurement data points. Results of numerical experiments are presented. The question of choosing the initial approximations is considered.

This article describes the development of a multifunctional geoinformation system RANDOM (Risk Assessment of Nature Detriment due to Oil spill Migration), realizing a multiprocessor calculation of probabilistic risk models to assess the negative impact of the oil spill on the biota of the North Caspian. The urgency of the problems associated with the development of oil fields in a very vulnerable shallow part of the Caspian Sea, where a major accident could have disastrous consequences. We describe a risk model of biota damage in the case of accidental environmental pollution. The model rests on the formalization of the notions of the accident, environmental pollution, biomass, and biota sensitivity, which depend on each other in the case of an accident. In addition, this article describes the development process from design to implementation. The system is designed on the basis of service-oriented architecture (SOA), which allows for easy, flexible integration of services, and access them via the Internet. Through the use of SOA, the system can be expanded and upgraded. In this approach, the services may be located on physically different servers. Tests have shown the benefit of using a supercomputer; it enables us to obtain a risk assessment for an adequate time. This system is designed for professionals in the field of ecology and mathematical modelling and subsoil oil fields on the continental shelf of the seas and oceans. RANDOM system as the final result of the decision of risk assessment tasks includes a series of calculation modules based on the methods of probability theory, computational mathematics, hydrodynamics, oil chemistry, marine biology, mathematical modelling, and geoinformatics.

In this paper, we consider the continuation problem for the Helmholtz equation. The main result is a conditional stability estimate for a solution to the considered problem. The estimate shows that the closer solution to the surface is more stable. The results of the numerical experiments confirm this conclusion.

Verhulst limited growth model with unknown parameters of growth is considered. These parameters are defined by discrete experiment data. This inverse problem is solved with using gradient method with interpolation of data and without it. Approximation of the delta-function is used for the latter case. As an example the bacteria population E.coli is considered.

We consider an ill-posed initial boundary value problem for the Helmholtz equation. This problem is reduced to the inverse continuation problem for the Helmholtz equation. We prove the well-posedness of the direct problem and obtain a stability estimate of its solution. We solve numerically the inverse problem using the Tikhonov regularization, Godunov approach, and the Landweber iteration. Comparative analysis of these methods is presented.