Pattern Recognition Tasks Research Articles

Syntactic Pattern Recognition in Computer Vision

Using techniques derived from the syntactic methods for visual pattern recognition is not new and was much explored in the area called syntactical or structural pattern recognition. Syntactic methods have been useful because they are intuitively simple to understand and have transparent, interpretable, and elegant representations. Their capacity to represent patterns in a semantic, hierarchical, compositional, spatial, and temporal way have made them very popular in the research community. In this article, we try to give an overview of how syntactic methods have been employed for computer vision tasks. We conduct a systematic literature review to survey the most relevant studies that use syntactic methods for pattern recognition tasks in images and videos. Our search returned 597 papers, of which 71 papers were selected for analysis. The results indicated that in most of the studies surveyed, the syntactic methods were used as a high-level structure that makes the hierarchical or semantic relationship among objects or actions to perform the most diverse tasks.

Parallel Computation of Discrete Orthogonal Moment on Block Represented Images Using OpenMP

Herein, a parallel implementation of Discrete Orthogonal moments on block represented images is investigated. Moments and moment functions have been used widely as features for image analysis and pattern recognition tasks. The main disadvantage of all moment sets, is the high computational cost which is increased as higher-order moments are involved in the computations. In image block representation (IBR) the image is represented by homogeneous areas which are called blocks. The IBR allows moment computation with zero computational error for binary images, low computational error for gray images, low computational complexity, while can achieve high processing rates. The results from parallel implementation on a multicore computer using OpenMP, exhibit significant performance.

METHOD OF SPECTRAL CLUSTERING OF PAYMENTS AND RAW MATERIALS SUPPLY FOR THE COMPLIANCE AUDIT PLANNING

Context. The analytical procedures used in the audit are currently based on data mining techniques. The work solves the problem of increasing the efficiency and effectiveness of analytical audit procedures by clustering based on spectral decomposition. The object of the research is the process of auditing the compliance of payment and supply sequences for raw materials.
 Objective. The aim of the work is to increase the effectiveness and efficiency of the audit due to the method of spectral clustering of sequences of payment and supply of raw materials while automating procedures for checking their compliance.
 Method. The vectors of features are generated for the objects of the sequences of payment and supply of raw materials, which are then used in the proposed method. The created method improves the traditional spectral clustering method by automatically determining the number of clusters based on the explained and sample variance rule; automatic determination of the scale parameter based on local scaling (the rule of K-nearest neighbors is used); resistance to noise and random outliers by replacing the k-means method with a modified PAM method, i.e. replacing centroid clustering with medoid clustering. As in the traditional approach, the data can be sparse, and the clusters can have different shapes and sizes. The characteristics of evaluating the quality of spectral clustering are selected.
 Results. The proposed spectral clustering method was implemented in the MATLAB package. The results obtained made it possible to study the dependence of the parameter values on the quality of clustering.
 Conclusions. The experiments carried out have confirmed the efficiency of the proposed method and allow us to recommend it for practical use in solving audit problems. Prospects for further research may lie in the creation of intelligent parallel and distributed computer systems for general and special purposes, which use the proposed method for segmentation, machine learning and pattern recognition tasks.

Low‐Power Self‐Rectifying Memristive Artificial Neural Network for Near Internet‐of‐Things Sensor Computing

AbstractFrequent data transfers between Internet‐of‐Things (IoT) sensors and cloud servers consume energy and lead to latency—a bottleneck for ubiquitous computing. To reduce the need for such enormous data transfers, the combined function of IoT sensors and near‐sensor artificial neural networks can process data properly before they are transferred to cloud servers. Herein, energy‐efficient memristor crossbar arrays are demonstrated for image recognition tasks that are potentially adopted for IoT sensors. The adoption of the selector‐free memristor device with a self‐rectifying function allows for simple stacking of metal–dielectric–metal layer, thus significantly simplifying the fabrication process while achieving low‐current operation (<10 µA in microdevice). Area‐dependent resistive switching characteristics and the incorporation of interface effects reveal the role of the switching and rectifying phenomena in such devices. Finally, the Modified National Institute of Standards and Technology pattern recognition task is demonstrated with 32 × 32 memristor crossbar arrays combining a SPICE simulation. Therefore, it is expected that self‐rectifying memristor arrays can pave the way for the development of more intelligent IoT sensors.

DeepTCR is a deep learning framework for revealing sequence concepts within T-cell repertoires

Deep learning algorithms have been utilized to achieve enhanced performance in pattern-recognition tasks. The ability to learn complex patterns in data has tremendous implications in immunogenomics. T-cell receptor (TCR) sequencing assesses the diversity of the adaptive immune system and allows for modeling its sequence determinants of antigenicity. We present DeepTCR, a suite of unsupervised and supervised deep learning methods able to model highly complex TCR sequencing data by learning a joint representation of a TCR by its CDR3 sequences and V/D/J gene usage. We demonstrate the utility of deep learning to provide an improved ‘featurization’ of the TCR across multiple human and murine datasets, including improved classification of antigen-specific TCRs and extraction of antigen-specific TCRs from noisy single-cell RNA-Seq and T-cell culture-based assays. Our results highlight the flexibility and capacity for deep neural networks to extract meaningful information from complex immunogenomic data for both descriptive and predictive purposes.

A Rapid Spiking Neural Network Approach With an Application on Hand Gesture Recognition

The spiking neural network (SNN) is considered to be the third generation of neural networks featured by its low power consumption and high computing capability, which has great application potential in robotics. However, the present SNN has two limitations: 1) the neuron’s spike firing time is calculated based on the iterative approach, which dramatically slows down the calculation rate of the SNN and 2) the existing learning algorithm is more suitable for the single-layer structure, which can hardly train the network with “deep structure.” To this end, this paper proposes a novel spike firing time search algorithm that can narrow the search interval. In addition, a pretrained subnet SNN is designed, which makes the SNN have more hidden layers. This setting of the SNN can effectively improve its performance in pattern recognition tasks. Furthermore, by using the surface electromyography signal (sEMG), the proposed SNN is used to recognize the hand gestures. The experimental results show that: 1) the spike firing time search algorithm can significantly increase the forward propagation rate of the SNN and 2) the proposed SNN can reach a satisfactory recognition accuracy ratio 97.4%, which is 0.9% higher than that of the fully connected SNN.

Deep Learning for EMG-based Human-Machine Interaction: A Review

Electromyography (EMG) has already been broadly used in human-machine interaction (HMI) applications. Determining how to decode the information inside EMG signals robustly and accurately is a key problem for which we urgently need a solution. Recently, many EMG pattern recognition tasks have been addressed using deep learning methods. In this paper, we analyze recent papers and present a literature review describing the role that deep learning plays in EMG-based HMI. An overview of typical network structures and processing schemes will be provided. Recent progress in typical tasks such as movement classification, joint angle prediction, and force/torque estimation will be introduced. New issues, including multimodal sensing, inter-subject/inter-session, and robustness toward disturbances will be discussed. We attempt to provide a comprehensive analysis of current research by discussing the advantages, challenges, and opportunities brought by deep learning. We hope that deep learning can aid in eliminating factors that hinder the development of EMG-based HMI systems. Furthermore, possible future directions will be presented to pave the way for future research.

Neuromorphic Binarized Polariton Networks.

The rapid development of artificial neural networks and applied artificial intelligence has led to many applications. However, current software implementation of neural networks is severely limited in terms of performance and energy efficiency. It is believed that further progress requires the development of neuromorphic systems, in which hardware directly mimics the neuronal network structure of a human brain. Here, we propose theoretically and realize experimentally an optical network of nodes performing binary operations. The nonlinearity required for efficient computation is provided by semiconductor microcavities in the strong quantum light-matter coupling regime, which exhibit exciton–polariton interactions. We demonstrate the system performance against a pattern recognition task, obtaining accuracy on a par with state-of-the-art hardware implementations. Our work opens the way to ultrafast and energy-efficient neuromorphic systems taking advantage of ultrastrong optical nonlinearity of polaritons.

ACCURACY IMPROVING OF PRE-TRAINED NEURAL NETWORKS BY FINE TUNING

Methods of accuracy improving of pre-trained networks are discussed. Images of ships are input data for the networks. Networks are built and trained using Keras and TensorFlow machine learning libraries. Fine tuning of previously trained convoluted artificial neural networks for pattern recognition tasks is described. Fine tuning of VGG16 and VGG19 networks are done by using Keras Applications. The accuracy of VGG16 network with finetuning of the last convolution unit increased from 94.38% to 95.21%. An increase is only 0.83%. The accuracy of VGG19 network with fine-tuning of the last convolution unit increased from 92.97% to 96.39%, which is 3.42%.

Robust fuzzy rough set based dimensionality reduction for big multimedia data hashing and unsupervised generative learning

The amount of high dimensional data produced by visual sensors in the smart environments and by autonomous vehicles is increasing exponentially. In order to search and model this data for real-time applications, the dimensionality of the data should be reduced. In this paper, a novel dimensionality reduction algorithm based on fuzzy rough set theory, called Centralized Binary Mapping (CBM), is proposed. The fuzzy CBM kernel is used for extracting the central elements and the memory cells from the blocks of high dimensional data. The proposed applications of CBM in this paper include hashing and generative modelling of multimedia big data. The robustness of the proposed CBM based hashing algorithm is 10% higher than comparable methods. Furthermore, based on the CBM, a novel architecture for neural networks called Deep Root Dimensional Mapping (DRDM) is proposed. The DRDM is used for generative modelling of multimedia big data using a new autonomous vehicle visual navigation dataset as well as the standard datasets. The simulation results show that the proposed DRDM converges rapidly and the perceptual quality of the outputs at the same epoch is higher than generative adversarial networks. The proposed CBM can be used as a new data structures in various pattern recognition and machine learning tasks.

Energy-efficient event pattern recognition in wireless sensor networks using multilayer spiking neural networks

Motivated by the energy-efficient computation of the brain, an energy-efficient wireless sensor network infrastructure is built for solving the pattern recognition task. In this work, spatiotemporal patterns originate from the wireless sensor nodes that use pulse-based networking schemes like pulse position-coded packet data unit and spiked time-division multiple access. Pulse-based networking schemes use spatiotemporal patterns to encode information rather than using traditional energy-expensive packets. These spatiotemporal patterns are learned through biologically plausible learning rule of Tempotron which is a neuronal binary classifier that uses a gradient-descent supervised learning rule to adjust the weights of synaptic inputs. A single Tempotron-trained spiking neuron is shown to read out information encoded in the spike timings of synaptic inputs. However, due to the simplicity of the single Tempotron classifier, its performance is inherently limited. In this work, we devise the multilayer spiking neuron training rules for event pattern classification in distributed wireless sensor networks. We also show that the proposed architecture improves classification accuracy by a considerable amount as compared to a single Tempotron model’s performance.

ClimateNet: an expert-labeled open dataset and deep learning architecture for enabling high-precision analyses of extreme weather

Abstract. Identifying, detecting, and localizing extreme weather events is a crucial first step in understanding how they may vary under different climate change scenarios. Pattern recognition tasks such as classification, object detection, and segmentation (i.e., pixel-level classification) have remained challenging problems in the weather and climate sciences. While there exist many empirical heuristics for detecting extreme events, the disparities between the output of these different methods even for a single event are large and often difficult to reconcile. Given the success of deep learning (DL) in tackling similar problems in computer vision, we advocate a DL-based approach. DL, however, works best in the context of supervised learning – when labeled datasets are readily available. Reliable labeled training data for extreme weather and climate events is scarce. We create “ClimateNet” – an open, community-sourced human-expert-labeled curated dataset that captures tropical cyclones (TCs) and atmospheric rivers (ARs) in high-resolution climate model output from a simulation of a recent historical period. We use the curated ClimateNet dataset to train a state-of-the-art DL model for pixel-level identification – i.e., segmentation – of TCs and ARs. We then apply the trained DL model to historical and climate change scenarios simulated by the Community Atmospheric Model (CAM5.1) and show that the DL model accurately segments the data into TCs, ARs, or “the background” at a pixel level. Further, we show how the segmentation results can be used to conduct spatially and temporally precise analytics by quantifying distributions of extreme precipitation conditioned on event types (TC or AR) at regional scales. The key contribution of this work is that it paves the way for DL-based automated, high-fidelity, and highly precise analytics of climate data using a curated expert-labeled dataset – ClimateNet. ClimateNet and the DL-based segmentation method provide several unique capabilities: (i) they can be used to calculate a variety of TC and AR statistics at a fine-grained level; (ii) they can be applied to different climate scenarios and different datasets without tuning as they do not rely on threshold conditions; and (iii) the proposed DL method is suitable for rapidly analyzing large amounts of climate model output. While our study has been conducted for two important extreme weather patterns (TCs and ARs) in simulation datasets, we believe that this methodology can be applied to a much broader class of patterns and applied to observational and reanalysis data products via transfer learning.

A Theoretical Foundation for Syntactico-Semantic Pattern Recognition

Conventionally syntactic pattern recognition tasks have been driven by grammars defining a syntactic structure. Syntactic Pattern recognition tasks were primarily relying on the ability of parsing algorithms to recognize the patterns in the input data. These algorithms essentially were dependent on the syntactic grammars defining the patterns. Context free grammars, a type of grammars have been particularly well studied for pattern recognition tasks to be solved by computer efficiently. Some of the key pattern recognition tasks had applications in Natural Language Processing (NLP). Though context free grammars are well suited for capturing rigid patterns and unambiguous patterns, there was a need to encapsulate the uncertainty aspects involved in some pattern recognition processes. Probabilistic context free grammars can well handle the need to capture uncertainty in the processes but not in a true sense they are able to capture the uncertainty associated with the semantic context governing the domain in which the pattern recognition processes are being attempted at. The paper formally puts forth an approach for syntactico-semantic pattern recognition. The syntactico-semantic pattern recognition attempts to capture the semantic context and the uncertainties involved thereof along with probabilistic reasoning. The approach consists of integration mapping between probabilistic context free grammar (PCFG) and Multi Entity Bayesian network (MEBN), a first-order logic for modeling probabilistic knowledge bases. Additionally, the paper outlines a modified version of the CYK parser algorithm for the defined mapping between PCFG and MEBN with a method to ensure the properness and consistency of such PCFG along with its key application, disambiguation of PP (Prepositional Phrase) attachment. The theoretical foundation proposed has been validated by a proof-of-concept implementation of the modified CYK algorithm for syntactico-semantic reasoning in Java with promising ability to disambiguate PP attachment uses cases of New York Times and Wikipedia corpus dataset samples.

Исследование методов выделения границ в задаче распознавания объектов на изображениях с низкой контрастностью

The article discusses edge detection methods separately and combinations of edge detection filters with antialiasing filters in the task of pattern recognition on images with low contrast. Sobel, Canny, Otsu and thresholding filters are considered as edge detection methods. Median and Gaussian filters are considered as smoothing filters. The performance of the filters is assessed using the peak signal-to-noise ratio (PSNR) and the structural similarity index (SSIM).

The construction of local convex hull on the task of pattern recognition

In this research paper analyzes and explores the problems of pattern recognition methods and biological object identification algorithms. We encountered two different problems in the process. The first is the modernization of the pattern recognition algorithm; the second is associated with the identification of a biological object. The algorithm problem was solved by constructing local convex hull during modernization. In the second, the Orthoptera insect group was selected as a biological object on the recommendation of experts. Orthoptera is a taxonomic insect division that includes locusts and other winged insects. Approaches to determining the characteristics of orthoptera types and preparing their samples, as well as the formation of identification models, are recommended. The algorithm for calculating estimates (ACE) was chosen when identifying the choice of orthoptera. The ACE based on the principle of partial precedence. Approaches to the formation of a training-test selection for the choice of orthopters are proposed. As mentioned above, in this work is proposed to develop an information recognition system designed to solve the problems associated with the identification of Orthoptera Insecta, as well as for the experimental study of the specie of the locust family.

Effects of Data Augmentation Method Borderline-SMOTE on Emotion Recognition of EEG Signals Based on Convolutional Neural Network

In recent years, with the continuous development of artificial intelligence and brain-computer interface technology, emotion recognition based on physiological signals, especially electroencephalogram signals, has become a popular research topic and attracted wide attention. However, the imbalance of the data sets themselves, affective features’ extraction from electroencephalogram signals, and the design of classifiers with excellent performance, pose a great challenge to the subject. Motivated by the outstanding performance of deep learning approaches in pattern recognition tasks, we propose a method based on convolutional neural network with data augmentation method Borderline-synthetic minority oversampling technique. First, we obtain 32-channel electroencephalogram signals from DEAP data set, which is the standard data set of emotion recognition. Then, after data pre-processing, we extract features in frequency domain and data augmentation based on the data augmentation algorithm above for getting more balanced data. Finally, we train a one dimensional convolutional neural network for three classification on two emotional dimensions valence and arousal. Meanwhile, the proposed method is compared with some traditional machine learning methods and some existing methods by other researchers, which is proved to be effective in emotion recognition, and the average accuracy rate of 32 subjects on valence and arousal are 97.47% and 97.76% respectively. Compared with other existing methods, the performance of the proposed method with data augmentation algorithm Borderline-SMOTE shows its advantage in affective emotional recognition than that without Borderline-SMOTE.

Адаптивный алгоритм снижения уровня импульсного шума на изображениях с камер видеонаблюдения

An optical signal is usually converted into an electrical one by using photosensitive matrices with a large number of discrete elements based on charge-coupled device (CCD) technology or CMOS technology. One of the disadvantages of CCD and CMOS technologies is the impulse conversion noise that appears on digitized images, impairing visual perception and significantly reducing the likelihood of correct identification in pattern recognition tasks. Traditionally, impulse noise is removed from images using median filters with a fixed aperture within each iteration of full-format processing. However, such filters reduce the sharpness of the reconstructed image at high noise levels or insufficiently suppress the interference under the same noise conditions. These setbacks call for a need to develop an adaptive median filtering algorithm, which would produce a reconstructed image as a joint result of processing with median filters with different apertures. The essence of this algorithm is to select image areas with different noise levels and process these areas with filters with different apertures. As an objective criterion for assessing the efficiency of the proposed filtering algorithm, the authors used the criterion of the maximum correlation coefficient between noise-free and non-noisy images at various values of the noise variance. The mathematical modeling performed in this study allowed finding that with an increase in the impulse noise variance, the gain of the adaptive median filtering algorithm increases exponentially, in comparison with the algorithms using the filters with a fixed aperture value. The proposed algorithm can be used for pre-preprocessing images intended for recognition by machine vision systems, scanning text, and improving subjective image characteristics, such as sharpness and contrast.

Fusing Multisource Data to Estimate the Effects of Urbanization, Sea Level Rise, and Hurricane Impacts on Long-Term Wetland Change Dynamics

Wetlands are endangered ecosystems that provide vital habitats for flora and fauna worldwide. They serve as water and carbon storage units regulating the global climate and water cycle, and act as natural barriers against storm-surge among other benefits. Long-term analyses are crucial to identify wetland cover change and support wetland protection/restoration programs. However, such analyses deal with insufficient validation data that limit land cover classification and pattern recognition tasks. Here, we analyze wetland dynamics associated with urbanization, sea level rise, and hurricane impacts in the Mobile Bay watershed, AL since 1984. For this, we develop a land cover classification model with convolutional neural networks (CNNs) and data fusion (DF) framework. The classification model achieves the highest overall accuracy (0.93), and f1-scores in woody (0.90) and emergent wetland class (0.99) when those datasets are fused in the framework. Long-term trends indicate that the wetland area is decreasing at a rate of –1106 m2/yr with sharp fluctuations exacerbated by hurricane impacts. We further discuss the effects of DF alternatives on classification accuracy, and show that the CNN & DF framework outperforms machine/deep learning models trained only with single input datasets.

Impact of Deep Learning on Transfer Learning : A Review

Transfer learning and deep learning approaches have been utilised in several real-world applications and hierarchical systems for pattern recognition and classification tasks. However, in few of the real-world machine learning situations, this presumption does not sustain since there are instances where training data is costly or tough to gather and there is continually a necessity to produce high-performance learners competent with more easily attained data from diverse fields. The objective of this review is to determine more abstract qualities at the greater levels of the representation, by utilising deep learning to detach the variables in the outcomes, formally outline transfer learning, provide information on present solutions, and appraise applications employed in diverse facets of transfer learning and deep learning. This can be attained by rigorous literature exploration and discussion on all presently accessible techniques and prospective research studies on transfer learning solutions of independent as well as big data scale. The conclusions of this study could be an effectual platform directed at prospective directions for devising new deep learning patterns for different applications and dealing with the challenges concerned.

Emulation of biphasic plasticity in retinal electrical synapses for light-adaptive pattern pre-processing.

Electrical synapses provide rapid, bidirectional communication in nervous systems, accomplishing tasks distinct from and complementary to chemical synapses. Here, we demonstrate an artificial electrical synapse based on second-order conductance transition (SOCT) in an Ag-based memristor for the first time. High-resolution transmission electron microscopy indicates that SOCT is mediated by the virtual silver electrode. Besides the conventional chemical synaptic behaviors, the biphasic plasticity of electrical synapses is well emulated by integrating the device with a photosensitive element to form an optical pre-processing unit (OPU), which contributes to the retinal neural circuitry and is adaptive to ambient illumination. By synergizing the OPU and spiking neural network (SNN), adaptive pattern recognition tasks are accomplished under different light and noise settings. This work not only contributes to the further completion of synaptic behaviour for hardware-level neuromorphic computing, but also potentially enables image pre-processing with light adaptation and noise suppression for adaptive visual recognition.