Recognition Errors Research Articles

Fiber grating loop ring-down strain sensors using overlap spectrum demodulation and machine learning algorithm

The fiber grating loop ring-down (FGLRD) can be used in strain sensors to reduce costs and enhance system sensitivity. Introducing the overlap spectrum demodulation technology in the FGLRD can improve the system's linearity and expand the measurement range. However, there is always strong amplified spontaneous emission (ASE) noise owing to erbium-doped fiber amplifiers (EDFAs) in the FGLRD structure. The noise increases the difficulty of detecting the output ring-down spectrum and reduces the detection accuracy. This study proposes a machine learning algorithm for data processing of chirped FGLRD strain sensors based on overlap spectrum demodulation. It uses a global average pooling one-dimensional convolutional neural network algorithm to construct a data recognition model. The experimental results show that the method can improve the recognition accuracy and reduce the recognition errors caused by ASE noise compared to the conventional ring-down spectral peak detection method. The model accuracy was 100 % for the training and test sets and 98 % for the validation set. The loss value was close to 0.06, and the recognition time was approximately 0.5 ms.

Image Noise Removal in Ultrasound Breast Images Based on Hybrid Deep Learning Technique.

Rapid improvements in ultrasound imaging technology have made it much more useful for screening and diagnosing breast problems. Local-speckle-noise destruction in ultrasound breast images may impair image quality and impact observation and diagnosis. It is crucial to remove localized noise from images. In the article, we have used the hybrid deep learning technique to remove local speckle noise from breast ultrasound images. The contrast of ultrasound breast images was first improved using logarithmic and exponential transforms, and then guided filter algorithms were used to enhance the details of the glandular ultrasound breast images. In order to finish the pre-processing of ultrasound breast images and enhance image clarity, spatial high-pass filtering algorithms were used to remove the extreme sharpening. In order to remove local speckle noise without sacrificing the image edges, edge-sensitive terms were eventually added to the Logical-Pool Recurrent Neural Network (LPRNN). The mean square error and false recognition rate both fell below 1.1% at the hundredth training iteration, showing that the LPRNN had been properly trained. Ultrasound images that have had local speckle noise destroyed had signal-to-noise ratios (SNRs) greater than 65 dB, peak SNR ratios larger than 70 dB, edge preservation index values greater than the experimental threshold of 0.48, and quick destruction times. The time required to destroy local speckle noise is low, edge information is preserved, and image features are brought into sharp focus.

Water quality assessment in coastal lagoons using improved back propagation deep neural network with PH level compensation sensors in smart environment

<p>In coastal lagoons, eutrophication is a significant ecological and environmental problem. The majority of the pollution and deterioration problems that coastal zones and their ecosystems face are due to human influence through urbanization and industrialization. In this research, evaluations of water quality were made throughout a few lagoons in India during both the wet and dry seasons to establish their suitability for harboring aquatic life and the impact of seasonality on their quality. In order to reduce the effect of environmental influences on sensor based on nanoparticle sensitive film, an enhanced deep Back Propagation (BP) neural system was developed. The incompatibility between high performance and low calculation time was typically found in today’s modern sensor compensation techniques. Moreover, higher recognition errors were always the outcome of weak self-learning abilities. To solve these problems, a 16-layer deep BP neural network model was developed after hyper parameter searching. In Addition, the stochastic gradient descent (SGD) method and the mini-batch method were employed to achieve a proper balance between model performance and training time difficulties. The enhanced deep BP neural network was a great compensation solution for the sensor used in complicated environments due to its self-learning and self-adaptability. The proposed method is named as Imp_BPNN_SGD is analyzed in terms of various parameters and found that it achieves 35.6% of RMSE, 21.6% of MAE, 19.6% of RAE and 97.2% of D2</p>

Defect identification in adhesive structures using multi-Feature fusion convolutional neural network

The interface-debonding defects of adhesive bonding structures may cause a reduction in bonding strength, which in turn affects the bonding quality of adhesive bonding samples. Hence, defect recognition in adhesive bonding structures is particularly important. In this study, a terahertz (THz) wave was used to analyze bonded structure samples, and a multi-feature fusion convolutional neural network (CNN) was used to identify the defect waveforms. The pooling method of the squeeze-and-excitation (SE) attention mechanism was optimized, defect feature weights were adaptively assigned, and feature fusion was conducted using automatic label net-works to segment the THz waveforms in the adhesive bonding area with fine granularity waveforms as an input to the multi-channel CNN. The results revealed that the speed of the THz waveform labeling with the automatic labeling network was 10 times higher than that with traditional methods, and the defect-recognition accuracy of the defect-recognition network constructed in this study was up to 99.28%. The F1-score was 99.73%, and the lowest pre-embedded defect recognition error rate of the generalization experiment samples was 0.27%.

A Financial Handwritten Digit Recognition Model Based on Artificial Intelligence

Handwritten digit recognition is a kind of image information classification problem through optical character analysis. Its essence is to represent the pixel points of digital images as gray values and then replace the pixel matrix with a numerical matrix. The computer can deal with numerical problems converted from handwritten Arabic numeral recognition problems through feature extraction and classification. With the rapid development of science and technology, this technology has dramatically reduced the cost of identification and human consumption, making identification more efficient and having a specific use value. However, the current handwritten digit recognition technology will cause problems such as abnormal recognition and recognition errors, reducing recognition accuracy. This will not only increase the cost of human recognition but also increase unnecessary risks. Based on the broad application prospect of handwritten digit recognition in finance, this paper focuses on the research and analysis of the handwritten digit recognition model for its insufficient accuracy, low performance, and other problems. Based on traditional data analysis, this paper adopts deep learning and control variable methods to conduct multiple groups of experiments to explore the impact of different parameters on the accuracy of experimental results. Summarize the best recognition accuracy and achieve the best model performance.

Recognition of Handwritten Words from Digital Writing Pad Using MMU-SNet

In this paper, Modified Multi-scale Segmentation Network (MMU-SNet) method is proposed for Tamil text recognition. Handwritten texts from digital writing pad notes are used for text recognition. Handwritten words recognition for texts written from digital writing pad through text file conversion are challenging due to stylus pressure, writing on glass frictionless surfaces, and being less skilled in short writing, alphabet size, style, carved symbols, and orientation angle variations. Stylus pressure on the pad changes the words in the Tamil language alphabet because the Tamil alphabets have a smaller number of lines, angles, curves, and bends. The small change in dots, curves, and bends in the Tamil alphabet leads to error in recognition and changes the meaning of the words because of wrong alphabet conversion. However, handwritten English word recognition and conversion of text files from a digital writing pad are performed through various algorithms such as Support Vector Machine (SVM), Kohonen Neural Network (KNN), and Convolutional Neural Network (CNN) for offline and online alphabet recognition. The proposed algorithms are compared with above algorithms for Tamil word recognition. The proposed MMU-SNet method has achieved good accuracy in predicting text, about 96.8% compared to other traditional CNN algorithms.

Распознавание выборочных распределений среди системы эталонов: метод ближайшего соседа

The value of the sample distribution identification error of a multidimensional discrete random variable among a library of reference patterns is studied, depending on the dimension of the random vector, the sample length and the distance between two reference distributions in the norms C and L1. It is shown that the recognition error in the L1 norm is significantly lower than in C. Reference distributions of n-grams for texts are considered as a practical application. It turned out that the accuracy of identification is mainly determined by the individual characteristics of the standards, and not by the distances between them. An algorithm has been developed to test the system of standards for recognition accuracy.

Stain-Detection and Anti-Stain Algorithms Based on Dual Detector Data Interchange in Image-Type Displacement Measurement Technology

Displacement measurement technology based on image recognition algorithms represents a novel high-performance technique. In such measurements, when there are stains contaminating the measuring light path, it causes recognition errors from the linear array image sensor on the grating, resulting in inaccurate measurement results. Therefore, this report proposes stain-detection and anti-stain algorithms based on image recognition. The principle of image-type displacement measurement technology is described, and the influence of stains on the displacement measurements is analyzed according to four criteria in order to judge whether there are stains present. Then, an anti-stain algorithm based on the data exchange of dual-image sensors is proposed. Finally, experimental verifications are carried out. The experimental results show that the proposed algorithm effectively identifies the errors in the displacement measurements, when the experimental device is contaminated, the developed algorithms can accurately self-correct for the stain’s interference. The results presented herein provide a theoretical basis for developing highly-reliable displacement measurement devices.

Minimising Biasing Word Errors for Contextual ASR With the Tree-Constrained Pointer Generator

Contextual knowledge is essential for reducing speech recognition errors on high-valued long-tail words. This paper proposes a novel tree-constrained pointer generator (TCPGen) component that enables end-to-end ASR models to bias towards a list of long-tail words obtained using external contextual information. With only a small overhead in memory use and computation cost, TCPGen can structure thousands of biasing words efficiently into a symbolic prefix-tree, and creates a neural shortcut between the tree and the final ASR output to facilitate the recognition of the biasing words. To enhance TCPGen, we further propose a novel minimum biasing word error (MBWE) loss that directly optimises biasing word errors during training, along with a biasing-word-driven language model discounting (BLMD) method during the test. All contextual ASR systems were evaluated on the public Librispeech audiobook corpus and the data from the dialogue state tracking challenges (DSTC) with the biasing lists extracted from the dialogue-system ontology. Consistent word error rate (WER) reductions were achieved with TCPGen, which were particularly significant on the biasing words with around 40% relative reductions in the recognition error rates. MBWE and BLMD further improved the effectiveness of TCPGen, and achieved more significant WER reductions on the biasing words. TCPGen also achieved zero-shot learning of words not in the audio training set with large WER reductions on the out-of-vocabulary words in the biasing list.

Error recognition in the Cantor cube

Abstract. Based on the notion of thin sets introduced recently by T. Banakh, Sz. Głąb, E. Jabłońska and J. Swaczyna we deliver a study of the infinite single-message transmission protocols. Such protocols are associated with a set of admissible messages (i.e. subsets of the Cantor cube ℤ 2 ω ). Using Banach-Mazur games we prove that all protocols detecting errors are Baire spaces and generic (in particular maximal) ones are not neither Borel nor meager. We also show that the Cantor cube can be decomposed to two thin sets which can be considered as the infinite counterpart of the parity bit. This result is related to so-called xor-sets defined by D. Niwiński and E. Kopczyński in 2014.

Research on Joint Extraction Method of Entity and Relation Triples Based on Hierarchical Cascade Labeling

As an important research field of artificial intelligence, knowledge graph develops rapidly, and triplet extraction is the key to the construction of a knowledge graph. The traditional pipeline extraction method will bring the error of entity recognition into the relationship extraction and affects the extraction effect. Besides, the traditional pipeline extraction method cannot solve the SEO (Single Entity Overlap) and EPO (Entity Pair Overlap) problems. Inspired by this, we compare the advantages and disadvantages of the mainstream methods of entity and relationship triples joint extraction, propose a new joint extraction method of entity relation triples based on a hierarchical cascade labeling model (named HCL model), and the HCL model is based on multi neural network cooperation. Further, we construct a balanced sampling Chinese dataset about the entity and relational triplet extraction which contains SEO and EPO. We carry out the experiments on the balanced data set, and the F1 value of the HCL model reaches 65.4% better than other baseline models.

Strawberry yield monitoring based on a convolutional neural network using high-resolution aerial orthoimages

This article presents the results of studies comparing the quality of work of two modern models of convolutional neural networks YOLOv7 and YOLOv8 used to monitor the yield of strawberries. To do this, we used the transfer method of machine learning models on a set of collected data consisting of four classes of development of generative formations of strawberry. As a result of the study, we obtained a data set that contained images of flowers, ovaries, mature and not mature berries. To ensure the balance of classes in the dataset, the Oversampling method was used, which included the generation of new images by applying various operations, such as resizing the image, normalizing brightness and contrast, converting images by rotating them by a certain angle and reflection, random noise addition, Gaussian blur. To collect data (images) in the field, a DJI Phantom 2 quadrocopter with a DJI Zenmuse Gimbal suspension and a GoPro HD HERO3 camera was used. To assess the quality of the YOLOv7 and YOLOv8 models when recognizing specified classes, well-known metrics were used that estimate the proportion of objects found that are really objects of a given class, such as Precision, Recall and mAP. Analysis of the results showed that the mAP metric for all classes of the YOLOv7 convolutional neural network model was 0,6, and the YOLOv8 model was 0,762. Analysis of the test sample images showed that the average absolute percentage error of image recognition of all classes by the YOLOv7 and YOLOv8 models was 9,2%. The most difficult to recognize was class the ovary of strawberries, the average absolute percentage error of which was 13,2%. In further studies, the use of high-resolution stereo cameras is recommended, which will further improve the accuracy of monitoring potential yields due to the possibility of determining the dimensional parameters of strawberry fruits and constructing 3D models of elevation maps using photogrammetry.

МЕТОДИ ПІДВИЩЕННЯ ЯКОСТІ ПЕРЕТВОРЕННЯ МОВИ НА ТЕКСТ В СИСТЕМАХ БІОМЕТРИЧНОЇ АУТЕНТИФІКАЦІЇ

The article discusses the methods and algorithms of speech-to-text conversion, modern open and commercial systems for creating systems, as well as the use of these technologies in the field of cyber security. It is proposed to create a high-quality speech-to-text conversion system. An analysis of the mathematical algorithms used to reduce the error rate, which makes it possible to create unique voice prints and increase protection against forgery, has been carried out. The structure of modern speech-to-text conversion systems is described. By changing datasets, parameters of hidden Markov models, a high-quality dictionary of phonemes, and the use of language models, there is an opportunity to reduce the percentage of errors in language recognition, as well as the use of a system for multilingualism such as "surzhyk". The mathematical methods of assessing the quality of the system of speech to text (WER), as well as various methods of calculation, which is important for their further improvement and optimization, are considered. The structure of modern systems is considered, namely, signal pre-processing, feature extraction, acoustic modeling, speech modeling, decoding, post-processing. For each of the stages, study vectors have been proposed that can reduce the error rate of the system as a whole. Reducing speech recognition errors and the ability to fake a voice is achieved using various methods: deep neural networks, hidden Markov models, Baum-Welch algorithm, N-gram models, models with attention, creation of a high-quality phonemes dictionary, dataset, and fillers. Speech-to-text conversion technology can be used in biometric authentication systems to detect and analyze the unique features of the user's voice. However, modern speech-to-text conversion systems for Ukrainian, Russian, and "surzhyk" need improvement in acoustic and language units. Scientific works, which are devoted to research and optimization of these systems for biometric authentication, do not fully cover these issues. This became the reason for further research in this direction, so this work aims to create a speech recognition system with a minimum error rate.

ДОСЛІДЖЕННЯ ТОЧНОСТІ ВИЗНАЧЕННЯ ЯКОСТІ РЕЧОВИН РАДІОХВИЛЬОВИМИ МЕТОДАМИ

Subject and Purpose. Accuracy of the material recognition using radio wave methods in the ultra-high frequency band for substanc- es with dielectric properties is the present paper concern. To estimate a total error arising in material measurements by the radio wave method and determine constituents of the error is the aim of the work. Methods and Methodology. The material recognition accuracy is estimated by the method of statistical analysis in terms of average statistical deviation and dispersion. Results. Advantages and disadvantages of the waveguide method in which a test material sample is placed inside a metal waveguide have been traced, suggesting an obvious drawback that the material recognition process of the sort is difficult to be automated. In the remote inspection procedure, the test material is in free space (e.g. on a conveyor) where it is illuminated with a microwave transmitting antenna. The receiving antenna is located on the other side of the test sample and transfers the received signal to the apparatus for determining material parameters. There, the attenuation coefficient is measured as the electromagnetic wave passes through the sample. The measurement results show a correlation dependence of the wave attenuation coefficient on the quality of the substance, enabling us to use frequency dependences of the material to reveal its unknown quality. The remote method makes it possible to automate the recognition of materials with dielectric properties. For these methods mentioned right above, random error values arising during the material recognition were estimated by the method of statistical analysis. Conclusions. The analysis of errors in the material recognition shows that the total error ranges from 7.28 to 12.74% with corresponding constituent errors including faults of today’s microwave measuring devices, inappropriate application of the method or unsuitable type of the structural model of the parameter determination, and errors in data calculations.

Challenges of Automatic Speech Recognition for medical interviews - research for Polish language

Automatic Speech Recognition (ASR) systems are quickly becoming a crucial element in supporting healthcare providers, improving the flow of information among medical teams, and enhancing the patient's experience. However, to be fully supportive, these ASR systems must meet certain requirements dictated by market realities: high accuracy of speech recognition and low rate of errors, the possibility of additional training the model, and the possibility of on-premise system installation. Therefore, the aim of this paper is to perform a comparative analysis of leading ASR systems available on the Polish market for the needs of conducting medical interviews. We selected three systems, Google ASR, Microsoft ASR, and Techmo ASR, and we compared their performance on a prepared data set of medical-related expressions spoken in Polish. The results of our analysis indicated that there are minor discrepancies in the accuracy of speech recognition between all three evaluated ASR systems, whereas only two ASR systems met the raised requirements, in both cases partially. Still, they all exhibited specific problems in recognising word endings or word boundaries. We were able to categorise such problems into: Misrecognitions, Quality Problems, and Word Boundaries, varying in their level of influence on the further speech recognition process. Our research findings are expected to provide valuable insights to a wide range of stakeholders facilitating the development of tailored speech recognition solutions that meet the specific needs of medical sector.

Semi-Supervised SAR ATR via Epoch- and Uncertainty-Aware Pseudo-Label Exploitation

Automatic target recognition (ATR) is a significant application scenario for Synthetic aperture radar (SAR) image interpretation. In recent years, deep-learning based SAR ATR approaches have made great progress, with the require of large amounts of labeled data for network training. However, the labeled SAR images are scarce, and the annotation of SAR images is expensive and time-consuming. In this paper, a novel semi-supervised learning (SSL) framework is proposed for SAR ATR, which effectively alleviates the need of labeled samples for network training, and allows for excavating the intrinsic semantic relationship information between samples belonging to different categories, greatly improving recognition performance. This method develops an epoch- and uncertainty-aware pseudo-label selection (EUAPS) mechanism, which takes advantage of the underutilized consistency between training epochs, and introduces the uncertainty estimates. EUAPS can select pseudo-label samples with high-confidence, allowing the network to perform well even when labels are extremely scarce. Furthermore, we propose a novel loss function, group relationship consistency loss (GRCL) that explicitly enforces the consistency of relations between different samples under perturbations. GRCL facilitates the network to learn rich sample relationship information and more discriminative features in sample group level, effectively reducing the recognition error rate. Extensive experiments were carried out on three public benchmark SAR datasets. And the experimental results illustrate that the proposed approach achieves the accuracy of 98.58% when 10 labeled samples per class and 88.65% when only 5 labeled samples per class respectively, outperforming the state-of-the-art SSL methods and demonstrating the superiority.

MONITORING OF THE EDUCATIONAL PROCESS USING ARTIFICIAL INTELLIGENCE METHODS

Considered the problem of increasing the effectiveness of the educational process due to the introduction of automatic control of attendance in the classroom using face recognition and additional information for the collection and further analysis of the received data. Algorithms and methods used in modern Facial Recognition Attendance System are studied. An intelligent system for monitoring the educational process and its analysis is proposed. Structural and functional schemes of the system, databases, software were developed, testing was carried out. During attendance monitoring, the webcam captures an image of the face of the participant of the educational process from the video stream, then the computer automatically creates a vector of facial features, which is compared with the vectors of facial features, pre-entered images and recorded in the relevant database. Vectors with 68 features are used for face recognition. In the development of the software, the tools of the OpenCV library and the Python programming language were used. With several successful comparisons, the person's data is identified: name and status (student or teacher), the current date and time are recorded in an Excel file. Operational data of the system is displayed on the monitor screen, which allows you to correct recognition errors if necessary. The system allows you to automatically keep a log of class attendance, create reports, analyze data to provide recommendations for improving the schedule and order of lessons. The object-relational database with open source code PostgreSQL is used for data storage. The grouping of the system's software code is carried out using the Django web application. The system user has the opportunity to create a personal account and create reports according to his requirements. The system was tested on the example of a group of 15 students and showed satisfactory results.

Review of Optical Character Recognition for Power System Image Based on Artificial Intelligence Algorithm

Optical Character Recognition (OCR) refers to a technology that uses image processing technology and character recognition algorithms to identify characters on an image. This paper is a deep study on the recognition effect of OCR based on Artificial Intelligence (AI) algorithms, in which the different AI algorithms for OCR analysis are classified and reviewed. Firstly, the mechanisms and characteristics of artificial neural network-based OCR are summarized. Secondly, this paper explores machine learning-based OCR, and draws the conclusion that the algorithms available for this form of OCR are still in their infancy, with low generalization and fixed recognition errors, albeit with better recognition effect and higher recognition accuracy. Finally, this paper explores several of the latest algorithms such as deep learning and pattern recognition algorithms. This paper concludes that OCR requires algorithms with higher recognition accuracy.

A Method for Eliminating Pig Face Recognition Errors Caused by Too Short Pig Growth Cycle

In the process of modern large-scale pig breeding, it is necessary to distinguish the identity of each pig and real-time detect its health status, weight change, dietary status, and other parameters. Traditional methods waste a lot of resources, while the high quality of pork cannot be effectively guaranteed. This project is based on convolutional neural networks to design and develop a pig face recognition system. This system uses an overhead camera suspended above the pig house to monitor the pig house for 24 hours and identify and track each pig. Due to the rapid growth cycle of the pig, the facial image information changes rapidly, which has a significant impact on the pig face recognition model. The acquisition camera module is designed to correct monitoring and tracking data. The acquisition camera is installed in the necessary place of the pig every day to collect real-time information and upload the collected information to the server. By comparing the data of individual pigs at different developmental stages with the server, the identification information of individual pigs is determined, and tracking data is corrected in a timely manner. At the same time, the monitoring and identification screen is displayed on the screen, and behavioral information parameters are recorded to facilitate the information management and breeding of the farm.

Facial recognition error leads to wrongful arrest of Black American

Facial recognition error leads to wrongful arrest of Black American