Search

Pattern recognition of optical label is utilized not only in packet switches but also in the field of data security due to its wide range of applications. In particular, pattern recognition using waveguide devices is highly valued because it eliminates the need for encoder processing. However, existing methods suffer from the problem of degraded contrast ratio because half of the input power are emitted from non‐target waveguides. In this article, a method is proposed to improve the contrast ratio employing a 4×$\times$4 multi‐mode interference (MMI) coupler with multiple input signals. The theoretical results reveal that the proposed method improves the contrast ratio from 3.0 dB to 9.5 dB and ∞$\infty$ dB for the three‐input and four‐input cases, respectively, for quadrature phase‐shift keying (QPSK)‐modulated signal. Furthermore, numerical simulation was performed through the three dimension finite‐difference time‐domain (3D‐FDTD) method, and the proposed scheme successfully discriminated each QPSK pattern. The power value |E|2$|E|^2$ at all ports closely matches the theoretical value, and the ratios of 9.5 dB and >$>$30.0 dB are numerically obtained. [ABSTRACT FROM AUTHOR]

AbstractDriven by the need to in situ monitoring of the endogenous substances in biological matrices, a simple, rapid and sensitive optical nanoarray system has been designed for determination and visual discrimination of four important catecholamine neurotransmitters: dopamine (DA), epinephrine (EP), norepinephrine (NE) and levodopa (LD). To fabricate this platform, spherical citrate-capped gold nanoparticles (AuNPs) in the size range of 13–42 nm have been employed as sensing elements. For each catecholamine different responses toward the AuNPs was obtained indicating unique aggregation kinetics. When the sensor array exposed to different catecholamines (CAs), both the color and UV–Vis absorbance spectra changed due to the interaction of analytes and AuNPs, creating a unique fingerprint-like pattern for each analyte which can be visually detected by naked eye. To analyze response profiles, we employed unsupervised chemometric algorithms including hierarchical cluster analysis (HCA) and principal component analysis (PCA) and satisfactory class separations were achieved. Furthermore, the sensitive RGB based color difference patterns were generated providing easily monitored visual output. Based on this sensor array, the set of four catecholamine neurotransmitters have been correctly identified and classified in the spiked real urine samples. The target analytes have been well distinguished at low concentrations of 0.3 − 4 μM DA, 0.8 − 10 μM EP, 0.8 − 8 μM NE, and 0.3 − 10 μM LD using the proposed “chemical nose” strategy. We believe that the established optical sensor array can be potentially employed as an efficient discrimination tool for diagnostic and therapeutic applications in the future. [ABSTRACT FROM AUTHOR]

Traditional preprocessing methods have problems of low stability and poor universality. To solve these problems, this article conducted effective research on parameter tuning of image preprocessing methods based on the Ant Colony Particle Swarm Optimization (ACPSO) algorithm. This article used peak signal-to-noise ratio (PSNR) and structural similarity index measure (SSIM) as fitness functions. The particle dimension is defined as 2, which represents the size of the median filter and the size of the histogram equalization window. After the iteration, the results of ant colony algorithm and particle swarm optimization (PSO) algorithm were compared, and the parameter with the highest fitness function value was selected as the final preprocessing parameter. The image was preprocessed using the determined optimal parameters. The results showed that the average PSNR and SSIM values of the ACPSO preprocessed images were 5.58 and 0.08 higher than those of traditional preprocessing methods, and the subjective visual evaluation score was also higher. The Otsu's binarization method was used for segmentation. The method of feature extraction using Histogram of Orientated Gradients (HOG) and Local Binary Patterns (LBP), as well as the recognition model using Convolutional Neural Networks (CNN) and Support Vector Machines (SVM), was more stable than traditional preprocessing methods. On the other two datasets, the recognition results of ACPSO preprocessed images performed better and showed better universality compared to traditional preprocessing methods. ACPSO algorithm, as a hybrid swarm intelligence algorithm, can effectively play an effective role in optical pattern recognition image preprocessing, solving the problems of low stability and poor universality of traditional preprocessing methods. [ABSTRACT FROM AUTHOR]

Optical pattern recognition (OPR) has the advantage of single intensity detection ability for low‐cost terahertz (THz) systems of imaging or security checks. However, conventional 4f‐system‐based OPR is limited by the paraxial approximation and bulky device volumes for THz applications. Here, a full‐diffraction‐based 4f‐less OPR method is proposed using a single complex‐amplitude‐modulating metasurface, which is valid for systems with large Fresnel numbers. Moreover, a laser‐induced graphene technique is applied for processing the device. A 15 mm × 15 mm metasurface can be fabricated by one‐step laser writing in 34 s, indicating the potential of the proposed method in developing THz OPR systems with miniaturization, fast fabrication, and low‐cost. [ABSTRACT FROM AUTHOR]

Many coal mining areas overlapped with agricultural land in the world. However, surface subsidence and waterlogging brought on by coal mining inexorably harm the agricultural land. Soil moisture can reflect the variation of mining impact significantly according to the geographic features of the high underground water table, which shows significant spatial gradient variation within the subsidence basin. Previous studies have used site survey and unmanned aerial vehicle (UAV), which are not applicable to large‐scale assessments of the ecological impacts and defining the scale of impacts. Here, we used the Google Earth Engine (GEE) platform and an optical trapezoid model (OPTRAM) to estimate the soil moisture distribution around 50 mining subsidence waterbody samples in eastern China. Based on the spatial variation in soil moisture, the impact boundary of coal mining was determined using buffer analysis and trend fitting. The results showed that: (1) The OPTRAM can well reflect the soil moisture distribution which shows a good correlation between estimated and measured values, with R = 0.83 and RMSE = 0.054. (2) The spatial trend fitting of soil moisture is suitable for evaluating mining impact, and the impact distances are in the range of 96–762 m, 80% of which are located at 128–436 m. We also find that 28% of subsidence waterbody buffer basins were heavily influenced by human activities, and the coal mining impact boundaries cannot be effectively identified. The study provides a new idea to quantify the ecological and agricultural impact boundary of coal mining, which could be applied to other similar regions across the globe. [ABSTRACT FROM AUTHOR]

The freestanding CSV file also has the advantage of being easily transportable by end-users who prefer to engage with Music DH as a spreadsheet instead of a website. (hereafter Music DH) (https://rutgersdh.github.io/musicdh/) represents the work of a collaborative project of the Digital Humanities Interest Group of the Music Library Association. The tools built into Minicomp/Wax helped the Music DH team process image data (specifically, a screenshot of each catalogue item) and assign metadata to those images in a standard way. An item-access statement has been manually added outside of the metadata block coming from the CSV to each of the four records with restricted content (in the website html it is denoted as [Access is limited, restricted or paywalled]). [Extracted from the article]

Automatic car license plate recognition (LPR) is widely used nowadays. It involves plate localization in the image, character segmentation and optical character recognition. In this paper, a set of descriptors of image segments (characters) was proposed as well as a technique of multi-stage classification of letters and digits using cascade of neural network and several parallel Random Forest or classification tree or rule list classifiers. The proposed solution was applied to automated recognition of number plates which are composed of capital Latin letters and Arabic numerals. The paper presents an analysis of the accuracy of the obtained classifiers. The time needed to build the classifier and the time needed to classify characters using it are also presented. [ABSTRACT FROM AUTHOR]

Health systems are human based systems comprising a high number of stakeholders who interact with each other in complex ways. Due to the stochastic nature of health systems and the complex dynamics and interactions of their inputs, processes and outputs; providers of health services require tools tat allow them to comprehend the said complexity and improve system performance. Intensive care units (ICUs) may fall short in treating patients due to insufficient capacity and patients may be transferred to other hospitals under certain circumstances, and waiting times can get longer. Efficient use and management of a scarce resource such as the ICU is of critical importance for trouble-free operation of a hospital. Capacity planning, in other words, determination of optical bed configuration based on current knowledge of future demand and capacity, is very difficult due to high uncertainty in both bed capacity and demand. The paper at hand proposes a simulation model to allow an easier decision-making process in the planning of bed capacity in ICU. The study focuses on improving the capacity planning of a level III university hospital in Turkey that specializes in education and research. The objective of the study is to determine optimal bed requirement by matching the resources of the ICU with demand. We provide forecasts of ICU capacity by evaluating the simulation models based on patient waiting times and bed quantity. Considered in the context of the vital importance of intensive care units and the uncertainties in their management, it is thought that the use of simulation modeling in the making of capacity decisions can prove useful for decision makers in assessing their resource allocation strategies under different regional and hospital conditions. [ABSTRACT FROM AUTHOR]

This pattern recognition method can effectively identify vibration signals collected by a phase-sensitive optical time-domain reflectometer (Φ-OTDR) and improve the accuracy of alarms. An alignment-free end-to-end multi-vibration event detection method based on Φ-OTDR is proposed, effectively detecting different vibration events in different frequency bands. The pulse accumulation and pulse cancellers determine the location of vibration events. The local differential detection method demodulates the vibration event time-domain variation signals. After the extraction of the signal time-frequency features by sliding window, the convolution neural network (CNN) further extracts the signal features. It analyzes the temporal relationship of each group of signal features using a bidirectional long short-term memory network (Bi-LSTM). Finally, the connectionist temporal classification (CTC) is used to label the unsegmented sequence data to achieve single detection of multiple vibration targets. Experiments show that using this method to process the collected 8563 data, containing 5 different frequency bands of multi-vibration acoustic sensing signal, the system F1 score is 99.49% with a single detection time of 2.2 ms. The highest frequency response is 1 kHz. It is available to quickly and efficiently identify multiple vibration signals when a single demodulated acoustic sensing signal contains multiple vibration events. [ABSTRACT FROM AUTHOR]

Recently found historical jewellery (Czech Republic) was subjected to detailed analyses to determine the gem inlays and the gold framework in which they are embedded. Such find fits into European jewellery archaeological artefacts containing similar stones from around the fifth century CE. The gemstones were analysed with optical microscopy and Raman micro-spectroscopy to determine their mineralogical characteristics and to find out typical structural-chemical differences based on which their provenance can be found. The results of this measurement discovered two main types of minerals from the garnet group, almandine and pyrope, where pyropes have been identified as Bohemian garnets thanks to the typical photoluminescence (PL) of chromium and vanadium impurities. The craftsmanship and processing of the goldsmith work were studied using X-ray computed tomography. Such a technique is an excellent contribution for detecting and visualising the internal parts hidden behind the placement of the stones. For this reason, 3D visualisation was used to describe and better understand all preserved parts. According to this object's very high cultural and historical value, the chosen methods are suitable for non-destructive study while proving to be essential to deepen our knowledge in archaeometric investigations focused on jewellery from Late Antiquity (with possible application to the jewellery from other periods). Based on the findings presented in this study, it is one of the world's oldest documented items using Bohemian garnets in jewellery. These findings underscore the tremendous economic and political importance of Bohemian garnet mineral resources altogether, with a second completely different type of garnet believed to have originated in distant India or Sri Lanka that was used in jewellery across Europe in Late Antiquity. [ABSTRACT FROM AUTHOR]

Motion capture is gaining significance in educational research. Different motion capture systems (MCSs) are used in educational research for different purposes and in different ways, which creates a diversity that is difficult to monitor. The aim of this article is to provide an overview of MCSs and their uses in educational research addressing primary and secondary school education. We conducted a systematic review focusing on the types of MCSs and the concrete systems being used, looking at how they are used, the purposes of their use, the aims and the research questions under investigation, the subjects/topics addressed in the studies, the locations/places of use, the potential benefits of using MCSs perceived by the researchers, and the underlying theories of movement. Based on n = 20 studies focusing on primary and secondary school education identified in our database search, we found that these studies mainly used optical MCSs. Furthermore, we identified three main purposes of their use: to directly support the learning process of students, as analysis tools, or for developing tutoring systems that support students in their learning. This paper gives insights into the use of MCSs in educational research and provides a springboard for further research using MCSs in educational research. [ABSTRACT FROM AUTHOR]

Copyright of Bulletin of the Chinese Ceramic Society is the property of Bulletin of the Chinese Ceramic Society Editorial Office and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use. This abstract may be abridged. No warranty is given about the accuracy of the copy. Users should refer to the original published version of the material for the full abstract. (Copyright applies to all Abstracts.)

Two types of transparent Y 2 O 3 ceramics without including large scattering sources such as residual pores, one with very high optical homogeneity (type A) and another one with slightly insufficient optical homogeneity (type B), are purposely prepared, and their optical properties are investigated and compared qualitatively and quantitatively. Type A ceramic exhibits transmittance characteristics with very low internal loss in the visible to infrared wavelength region, while type B ceramic is inferior in various optical performances especially in the short (visible) wavelength region. In type B ceramic, birefringence occurs due to optical inhomogeneity in the visible region, resulting in a decrease in the extinction ratio. Non-uniform refractive index distribution is also observed in the Schlieren image of type B ceramic, hence the laser beam quality through that material is degraded. This study proved the importance of optical homogeneity of transparent ceramics and clarified the problems in actual applications. [ABSTRACT FROM AUTHOR]

Get in-depth understanding of each part of visual pathway yields insights to conquer the challenges that classic computer vision is facing. Here, we first report the bioinspired striate cortex with binocular and orientation selective receptive field based on the crossbar array of self-powered memristors which is solution-processed monolithic all-perovskite system with each cross-point containing one CsFAPbI3 solar cell directly stacking on the CsPbBr2I memristor. The plasticity of self-powered memristor can be modulated by optical stimuli following triplet-STDP rules. Furthermore, plasticity of 3 × 3 flexible crossbar array of self-powered memristors has been successfully modulated based on generalized BCM learning rule for optical-encoded pattern recognition. Finally, we implemented artificial striate cortex with binocularity and orientation selectivity based on two simulated 9 × 9 self-powered memristors networks. The emulation of striate cortex with binocular and orientation selectivity will facilitate the brisk edge and corner detection for machine vision in the future applications. Designing efficient bio-inspired vision systems remains a challenge. Here, the authors report a bio-inspired striate visual cortex with binocular and orientation selective receptive field based on self-powered memristor to enable machine vision with brisk edge and corner detection in the future applications. [ABSTRACT FROM AUTHOR]

This study suggests a mechanism to help people with visual impairments navigate and move about. Nowadays, multiple services are possible at the same moment. Some of these will be discussed later in this essay. Nevertheless, no dependable and cost-effective alternative has yet been presented to replace the existing technology used by the visual impairments in their everyday tasks. This report begins with examining the problem, followed by an examination of the original problem resolution. Following that, it looked at some of the newest advancements and research in the field of helpful technologies. Finally, it suggests developing and implementing a system to assist the physically handicapped. A Raspberry microprocessor, cameras, batteries, eyeglasses, headphone, backup charger, and connections are all included in the suggested gadget. Use the camera to capture subjects. The phone's RCNN technique and deep learning modules will be used to finish computer vision and analysis. The ultimate output, on the other hand, is sent from the headgear to the vision impairer's hearing. The study offers methods as well as answers to the aforementioned issues. The research work may be applied to real-world scenarios and is suitable for people with visual impairments. For picture data processing, the gadget presented in this project employs an area accumulating neural net and a Microcontrollers. The process utilizes the Tesseract module from the Python language to do OCR (optical character recognition) and offers an output to the user. Later in this section, the prescribed tactics and effects are outlined. [ABSTRACT FROM AUTHOR]

Compared to inefficient, energy-intensive digital computing, optical analog computing has significant advantages in terms of computational speed and power consumption. Traditional optical analog computing is typically based on Fourier lenses and spatial frequency filters, which are bulky and difficult to integrate. These disadvantages can be overcome by replacing Fourier lenses and spatial frequency filters in traditional optical analog computing via a metasurface. We showcase a real-time compact optical cross-correlation analog computing based on a dielectric metasurface. An extra phase profile is introduced in the metasurface to approximate the Fresnel diffraction integral as the Fourier transform to realize a more compact scheme. The optical cross-correlation analog computing is derived by directly modulating the impinging optical field. A reference circular pattern of a circular target is highlighted recognized and identified by the operation. The recognition accuracy was analyzed by comparing the recognition peaks with the shifted center of the circle, showing that the operation is affected by the side flap of the other circle pattern in the background. Optical analog computing with a compact and powerful kernel may find potential solutions for bio-detection, pattern recognition, information retrieval, and electron tomography. • Analog computing can increase computational speed and decrease power consumption. • Metasurface is an alternative to space optics components to reduce size or volume. • A scheme for ultracompact optical cross-correlation analog computing is proposed. • The recognition of a circular target is achieved in theory and simulation. [ABSTRACT FROM AUTHOR]

• An attention-guided module, namely the Calibrated Augmented Attention Module (CAAM), is proposed for masked face recognition. • A recursive attention gate (RAG) is proposed that strengthens the layer-wise interactions to compute a joint attention response. • An augmented feature calibration block (AFCB) is proposed that incorporates cross-dimensional interactions for feature calibration. • Extensive experiments on various synthetic and realistic-masked face datasets clearly demonstrate the effectiveness of our method. Along with other aspects of daily life, the COVID-19 pandemic has a substantial impact on the performance of facial recognition (FR) systems installed in various locations for identity verification. To address this pivotal issue, we propose an attention-guided masked face recognition (MFR) method, named Calibrated Augmented Attention Module (CAAM), which consists of two core components: Recursive Attention Gate (RAG) and an Augmented Feature Calibration Block (AFCB). In the first stage, RAG guides the backbone network to pay attention to non-occluded face regions for feature learning by calibrating multi-layer features while progressively reducing the network's response to mask-occluded regions in a recursive manner. In the second stage, a dual-branch AFCB first augments the attention map generated by RAG to incorporate cross-dimensional interactions, which are then calibrated to build spatial and inter-channel dependencies across informative spatial locations for MFR. Experiments conducted on various masked face datasets validate the superior performance of CAAM. [ABSTRACT FROM AUTHOR]

Automatic license plate recognition (ALPR) is a critical technology for intelligent transportation systems. Most existing ALPR methods are focused on specific application scenarios. Although there are a few methods that focus on unconstrained scenarios, they are very time-consuming. In this work, we propose an efficient ALPR (EALPR) framework, where we can handle distorted license plates (LP) caused by perspective problems with high efficiency. We design a light LPD structure based on efficient object detection methods and use anchor-free strategies for LPD to alleviate the problem of expensive costs. Benefitting from these optimizations and a united framework structure, the proposed EALPR has real-time efficiency. We evaluate our method on five datasets and the results show that our method achieves state-of-the-art accuracy: 98.15% on OpenALPR(EU), 95.61% on OpenALPR(BR), 99.51% on AOLP(RP), 88.81% on SSIG, 79.41% on CD-HARD. Additionally, our method achieves an impressive speed of 74.9 FPS (Frames Per Second), outperforming existing approaches and demonstrating its efficiency. Our source code can be accessed at https://github.com/wechao18/Efficient-alpr-unconstrained. [ABSTRACT FROM AUTHOR]

• Review and categorize human action recognition algorithms for aerial data to offer insights into the current state of the art. • Provide a detailed discussion of aerial datasets related to human action recognition. • Offer future research directions by analyzing the challenges and limitations imposed by existing approaches. Human Action Recognition from Unmanned Aerial Vehicles is a dynamic research domain with significant benefits in scale, mobility, deployment, and covert observation. This paper offers a comprehensive review of state-of-the-art algorithms for human action recognition and provides a novel taxonomy that categorizes the reviewed methods into two broad categories: Localization based and Globalization based. These categories are defined by how actions are segmented from visual data and how their spatial and temporal structures are modeled. We examine these techniques, highlighting their strengths and limitations, and provide essential background on human action recognition, including fundamental concepts and challenges in aerial videos. Additionally, we discuss existing datasets, enabling a comparative analysis. This survey identifies gaps and suggests future research directions, serving as a catalyst for advancing human action recognition in aerial videos. To our knowledge, this is the first detailed review of this kind. [ABSTRACT FROM AUTHOR]

[Display omitted] • This paper aims to address the issue of slow testing speeds in deep graph convolutional models for skeleton-based action recognition by simplifying the network structure through knowledge distillation, resulting in a compact and lightweight student GNN. • To enhance the robustness and generalization capabilities of the model, we introduce a novel data augmentation mechanism that generates diverse action sequences while maintaining consistent behavior labels, providing a more comprehensive learning basis for the model. • Our model integrates three distinct knowledge learning paths: teacher networks, original datasets, and generated data. The combination of knowledge distillation and data augmentation enables the lightweight student network to surpass the large teacher network in terms of both performance and efficiency. • Experimental results demonstrate the effectiveness of our approach in the context of skeleton-based human action recognition, highlighting its potential to streamline state-of-the-art models while improving their performance. • This research provides a new perspective for skeleton-based action recognition, namely, combining knowledge distillation and data augmentation to achieve efficient and accurate behavior recognition, offering reference and inspiration for future related studies. Human action recognition, a pivotal topic in computer vision, is a highly complex and challenging task. It requires the analysis of not only spatial dependencies of targets but also temporal changes in these targets. In recent decades, the advancement of deep learning has led to the development of numerous action recognition methods based on deep neural networks. Given that the skeleton points of the human body can be treated as a graph structure, graph neural networks (GNNs) have emerged as an effective tool for modeling such data, garnering significant interest from researchers. This paper aims to address the issue of low test speed caused by over-complicated deep graph convolutional models. To achieve this, we compress the network structure using knowledge distillation from a teacher-student architecture, leading to a compact and lightweight student GNN. To enhance the model's robustness and generalization capabilities, we introduce a data augmentation mechanism that generates diverse action sequences while maintaining consistent behavior labels, thereby providing a more comprehensive learning basis for the model. The proposed model integrates three distinct knowledge learning paths: teacher networks, original datasets, and derived data. The fusion of knowledge distillation and data augmentation enables lightweight student networks to outperform their teacher networks in terms of both performance and efficiency. Experimental results demonstrate the efficacy of our approach in the context of skeleton-based human action recognition, highlighting its potential to simplify state-of-the-art models while enhancing their performance. [ABSTRACT FROM AUTHOR]

• The L e-2e nonconvex regularization is proposed to approximate the sparse prior of natural images. • A new restoration model for solving clear images is constructed. • An efficient algorithm is designed to optimize the proposed model. • The presented method outperforms the state-of-the-art methods on synthetic datasets and real-world images. This paper proposes a simple model for image deblurring with a new total variation regularization. Classically, the L 1-21 regularizer represents a difference of anisotropic (i.e. L 1) and isotropic (i.e. L 21) total variation, so we define a new regularization as L e-2e , which is the weighted difference of the mixed anisotropic (i.e. L 0 + L 1 = L e) and mixed isotropic (i.e. L 0 + L 21 = L 2e), and it is characterized by sparsity-promoting and robustness in image deblurring. Then, we merge the L 0 -gradient into the model for edge-preserving and detail-removing. The union of the L e-2e regularization and L 0 -gradient improves the performance of image deblurring and yields high-quality blur kernel estimates. Finally, we design a new solution format that alternately iterates the difference of convex algorithm, the split Bregman method, and the approach of half-quadratic splitting to optimize the proposed model. Experimental results on quantitative datasets and real-world images show that the proposed method can obtain results comparable to state-of-the-art works. [ABSTRACT FROM AUTHOR]

Gait recognition, an essential branch of biometric identification, uses walking patterns to identify individuals. Despite its effectiveness, gait recognition faces challenges such as vulnerability to changes in appearance due to factors like angles and clothing conditions. Recent progress in deep learning has greatly enhanced gait recognition, especially through methods like deep convolutional neural networks, which demonstrate impressive performance. However, current approaches often overlook the connection between coarse-grained and fine-grained features, thereby restricting their overall effectiveness. To address this limitation, we propose a new framework for gait recognition framework that combines deep-supervised fine-grained separation with coarse-grained feature learnability. Our framework includes the LFF module, which consists of the SSeg module for fine-grained information extraction and a mechanism for fusing coarse-grained features. Furthermore, we introduce the F-LCM module to extract local disparity features more effectively with learnable weights. Evaluation on CASIA-B and OU-MVLP datasets shows superior performance compared to classical networks. • Effective use of coarse- and fine-grained features of gait images. • Introducing learnable weights for better extraction of local variability features. • Adaptive fusion of features with training parameters. • Use deep supervised learning ideas. [ABSTRACT FROM AUTHOR]

A fundamental issue in skeleton-based action recognition is the extraction of useful features from skeleton joints. Unfortunately, the current state-of-the-art models for this task have a tendency to be overly complex and parameterized, which results in low model training and inference time efficiency for large-scale datasets. In this work, we develop a simple but yet an efficient baseline for skeleton-based Human Action Recognition (HAR). The architecture is based on adaptive GCNs (Graph Convolutional Networks) to capture the complex interconnections within skeletal structures automatically without the need of a predefined topology. The GCNs are followed and empowered with an attention mechanism to learn more informative representations. This paper reports interesting accuracy on a large-scale dataset NTU-RGB+D 60, 89.7% and 95.0% on respectively Cross-Subject, and Cross-View benchmarks. On NTU-RGB+D 120, 84.6% and 85.8% over Cross-Subject and Cross-Setup settings, respectively. This work provides an improvement of the existing model SGN (Semantic-Guided Neural Networks) when extracting more discriminant spatial and temporal features. [Display omitted] • Enhancing the existing model SGN's discriminant Spatio-Temporal feature extraction. • Our model is a lightweight network compared with the State-Of-The-Art (SOTA) models. • We report interesting accuracy compared with SOTA methods on large-scale dataset NTU. • Multi-input architecture provided complementary information about the skeleton data. [ABSTRACT FROM AUTHOR]

This paper proposed an efficient and novel end-to-end text detection and recognition framework called DiZNet. DiZNet is built upon a core representation using text detail maps and employs the classical lightweight ResNet18 as the backbone for the text detection and recognition algorithm model. The redesigned Text Attention Head (TAH) takes multiple shallow backbone features as input, effectively extracting pixel-level information of text in images and global text positional features. The extracted text features are integrated into the stackable Feature Pyramid Enhancement Fusion Module (FPEFM). Supervised with text detail map labels, which include boundary information and texture of important text, the model predicts text detail maps and fuses them into the text detection and recognition heads. Through end-to-end testing on publicly available natural scene text benchmark datasets, our approach demonstrates robust generalization capabilities and real-time detection speeds. Leveraging the advantages of text detail map representation, DiZNet achieves a good balance between precision and efficiency on challenging datasets. For example, DiZNet achieves 91.2% Precision and 85.9% F-measure at a speed of 38.4 FPS on Total-Text and 83.8% F-measure at a speed of 30.0 FPS on ICDAR2015, it attains 83.8% F-measure at 30.0 FPS. The code is publicly available at: https://github.com/DiZ-gogogo/DiZNet • Text Attention Head (TAH): To augment the model's representation capacity, we introduced TAH. This component efficiently extracts essential text features from shallow feature maps at varying resolutions derived from the ResNet18 backbone. These features are then integrated into the input for further enhancement. • Feature Pyramid Enhancement Fusion Module (FPEFM): A stackable module composed of a pyramid structure built using a feature pyramid and depth-wise separable convolutions. It integrates pixel-level text details and global positional features extracted by TAH, enhancing the model's feature representation capacity. The residual design of FPEFM and TAH strengthens the model's feature extraction, representation capabilities, and adaptive text feature extraction. • Text Detail Map: Harnessing the benefits of Text Detail Map representation, we introduced a Text Detail Head to predict Text Detail Maps. These predicted maps are integrated into both the text detection and recognition modules, markedly improving the model's robustness. • Overall Innovation Our work presents an efficient end-to-end algorithm for natural scene text detection and recognition grounded in Text Detail Maps, capturing both text texture and boundary information. The integration of TAH adeptly extracts pixel-level text details and global positional features, enhancing feature representation. The incorporation of residual connections further bolsters detection accuracy and speeds up post-processing inference. [ABSTRACT FROM AUTHOR]

• A weak surface defect detection method is proposed for production-line plastic bottles. • A multi-view imaging system is designed to capture surface images of plastic bottles. • A lightweight feature fusion YOLO network is developed. • The proposed method has wide potential applications. • The proposed method can accurately and quickly detect defects. The automatic optical detection (AOD) technique has been extensively applied in industrial manufacturing fields. However, in the manufacturing field of plastic bottles (PBs), the current AOD approaches often encounter the issues of 360-degree image acquisition, detection accuracy, and detection speed limitation due to the restriction of the imaging system, detection algorithm, and computational power. To address these issues, we propose a surface defect detection (SDD) method to identify weak and small defects in the production-line PBs. We design a multi-view imaging system containing four imaging units and four relative illumination devices to acquire the surface images of production-line PBs with a 360-degree angle of view. We develop a lightweight feature fusion (LFF) YOLO network to detect weak and small defects in the PB images. An experimental system is built to validate the proposed method. The practical experiment results illustrate that the proposed method realizes the detection of weak and small defects for the PBs with high accuracy and speed and performs better than the comparison state-of-the-art methods. The proposed method has wide applications in the three-dimensional object SDD, such as glass bottles, automobile parts, and so on. [ABSTRACT FROM AUTHOR]

The article titled "Retraction Note: Optical pattern recognition image preprocessing based on hybrid cluster intelligent algorithm" has been retracted by the publisher due to concerns about the peer-review process, inappropriate references, and the article's alignment with the journal's scope. The publisher, in consultation with the Editors-in-Chief, no longer has confidence in the results and conclusions of the article. It is important to note that the author disagrees with this retraction. Springer Nature, the publisher, remains neutral regarding jurisdictional claims and institutional affiliations. [Extracted from the article]

This study presents the design and optimization of an aerobics movement recognition system utilizing high-dimensional biotechnological data. Deep learning techniques are employed to achieve accurate classification and recognition of movement actions. Biosensing technology and wearable devices are used to collect real-time, multidimensional physiological signal data from key anatomical regions of athletes. The system is constructed using convolutional neural networks (CNNs) and Long Short-Term Memory. Model performance is optimized through parameter selection and strategies such as Xavier initialization, the cross-entropy loss function, and the Adam optimizer. The results indicate that Model C achieves an accuracy of 0.987, significantly outperforming the standalone CNNs (accuracy 0.975) and the recurrent neural network models (accuracy 0.965). Furthermore, it demonstrates notable efficiency in practical applications, with considerably reduced execution times of 10 s for data processing, 25 s for feature extraction, and 20 s for classification. This aerobics recognition system excels in performance and efficiency, supporting precise movement classification. [ABSTRACT FROM AUTHOR]

The advent of artificial intelligence has highly impacted the process of image processing and pattern recognition, hence influencing biomedical researchers to implement machine learning (ML)-assisted image-based analysis in the development of optical sensors. Optical detection devices provide the optimal platforms for image-based analysis as they exhibit signals that can be captured by cameras. Relying on ML for pattern recognition from the captured images can highly enhance the detection accuracy and pave the way toward developing point-of-care testing. In this review, we have aimed to present an overview based on image processing and how ML can be implemented in pattern recognition from images. Recent publications on the development of optical detection devices relying on ML-assisted image analysis in various fields of biomedical analysis have then been reviewed. • The implementation of image processing in optical sensing approaches. • The methods and advantages for machine learning-based image analysis. • The benefits of machine learning in image-based optical sensors. • Recent studies in biomedical sensing via machine learning-assisted image analysis. [ABSTRACT FROM AUTHOR]

For many years, optics has been employed in computing, although the major focus has been and remains to be on connecting parts of computers, for communications, or more fundamentally in systems that have some optical function or element (optical pattern recognition, etc.). Optical digital computers are still evolving; however, a variety of components that can eventually lead to true optical computers, such as optical logic gates, optical switches, neural networks, and spatial light modulators have previously been developed and are discussed in this paper. High-performance off-the-shelf computers can accurately simulate and construct more complicated photonic devices and systems. These advancements have developed under unusual circumstances: photonics is an emerging tool for the next generation of computing hardware, while recent advances in digital computers have empowered the design, modeling, and creation of a new class of photonic devices and systems with unparalleled challenges. Thus, the review of the status and perspectives shows that optical technology offers incredible developments in computational efficiency; however, only separately implemented optical operations are known so far, and the launch of the world's first commercial optical processing system was only recently announced. Most likely, the optical computer has not been put into mass production because there are still no good solutions for optical transistors, optical memory, and much more that acceptance to break the huge inertia of many proven technologies in electronics. [ABSTRACT FROM AUTHOR]

Background/Aims: Falling among the elderly is a major public health issue, especially with the advancing age of the baby boomers. The fall risk assessment tests for many lack a context that would bring them closer to everyday life. Thus, immersive virtual reality, which makes it possible to simulate everyday situations, could make it possible to strengthen the quality of the assessment of the risk of falls. However, it is necessary to understand how the use of a virtual reality device influences the motor control of elderly participants. If vestibular physiotherapists use VR to virtualize their tools, what impact would a visual simulation of movement have on motor control in a locomotor task, if this simulation were plausible? Methods: Sixty-two elders (70.8 ± 6.7 years old) completed a Timed Up and Go task under 3 conditions: real, virtual reality, and virtual reality with visual and sound movement information. The virtual reality task takes place in a train either stationary at a station or in uniform linear motion. The time and number of steps were recorded using video, and comparisons between conditions were made using Friedman's test. Results: The results show a significant increase in the time and number of steps in "virtual reality" condition compared to the "real" condition. They do not show significant differences between the 2 virtual conditions. Conclusion: The use of a running virtual train to provide plausible movement is particularly distinguished from vestibular physiotherapy applications with first a fixed visual support partially obscuring the optical flow. This visual aid coupled with the attention dedicated to the task inhibits the effect of the moving environment on locomotion. However, the visual optical flow will potentially have an effect in people with fear of falling. Virtual reality shows great potential for the simulation of realistic environments for the assessment of the risk of falls and opens up avenues for the development of tests. [ABSTRACT FROM AUTHOR]

The structure of five rare filigree spheres from a seventeenth century shipwreck was examined in order to unravel their condition, manufacturing process and function. This study focuses on the application of non-invasive imaging techniques: optical microscopy, X-radiography, X-ray micro-computed tomography and neutron computed tomography. A valuation of different aspects of the applied techniques was made, aiding stakeholders in decision-making on research and conservation. The combination of theory and scientific information was used to obtain an improved understanding of the manufacturing process and function of the filigree spheres. [ABSTRACT FROM AUTHOR]

With the wide use of Automatic Speech Recognition (ASR) in applications such as human machine interaction, simultaneous interpretation, audio transcription, and so on, its security protection becomes increasingly important. Although recent studies have brought to light the weaknesses of popular ASR systems that enable out-of-band signal attack, adversarial attack, and so on, and further proposed various remedies (signal smoothing, adversarial training, etc.), a systematic understanding of ASR security (both attacks and defenses) is still missing, especially on how realistic such threats are and how general existing protection could be. In this article, we present our systematization of knowledge for ASR security and provide a comprehensive taxonomy for existing work based on a modularized workflow. More importantly, we align the research in this domain with that on security in Image Recognition System (IRS), which has been extensively studied, using the domain knowledge in the latter to help understand where we stand in the former. Generally, both IRS and ASR are perceptual systems. Their similarities allow us to systematically study existing literature in ASR security based on the spectrum of attacks and defense solutions proposed for IRS, and pinpoint the directions of more advanced attacks and the directions potentially leading to more effective protection in ASR. In contrast, their differences, especially the complexity of ASR compared with IRS, help us learn unique challenges and opportunities in ASR security. Particularly, our experimental study shows that transfer attacks across ASR models are feasible, even in the absence of knowledge about models (even their types) and training data. [ABSTRACT FROM AUTHOR]

Optoelectronic synapses combine the functionalities of a non-volatile memory and photodetection in the same device, paving the path for the realization of artificial retina systems which can capture, pre-process, and identify images on the same platform. Graphene/Ta2O5/graphene phototransistor exhibits synapse characteristics when visible electromagnetic radiation of wavelength 405 nm illuminates the device. The photocurrent is retained after light withdrawal when positive gate voltage is applied to the device. The device exhibits distinct conductance states, modulated by different parameters of incident light, such as pulse width and number of pulses. The conductance state can be retained for 104 s, indicating long term potentiation (LTP), similar to biological synapses. By using optical and electrical pulses, the device shows optical potentiation and electrical LTD repeatably, implying their applicability in neural networks for pattern recognition. [ABSTRACT FROM AUTHOR]

Semantics and motion are two cues of essence for the success in video salient object detection. Most existing deep-learning based approaches extract semantic features by the use of only one fully convolutional network with simple stacked encoders. They simulate motion patterns of video objects with two consecutive frames being simultaneously fed into a convolutional LSTM network or a weights-sharing fully convolutional network. However, such approaches have the shortcomings of producing a coarse predicted saliency map or requiring significant computational overheads. In this paper, we present a novel approach with cascaded fully convolutional networks involving motion attention (abbreviated as CFCN-MA), to achieve real-time saliency detection in videos. Our key idea is to construct twofold fully convolutional networks in order to gain a saliency map from coarse to fine. We devise an optical flow-based motion attention mechanism to improve the prediction accuracy of the initial fully convolutional networks, using the popular FlowNet2-SD model that is efficient and effective for motion pattern recognition of distinctive objects in videos. This method can obtain a fine saliency map with a refined region of interest. Moreover, we propose a means for calculating attention-guided intersection-over-union loss (shortnamed as AIoU) to supervise the CFCN-MA model in learning a saliency map with both clear edge and complete structure. Our approach is evaluated on three popular benchmark datasets, namely DAVIS, ViSal and FBMS. Experimental results demonstrate that our method outperforms many state-of-the-art techniques while meeting the real-time demand at 27 fps. [ABSTRACT FROM AUTHOR]

This study aimed to clarify whether the pattern recognition involved in scoring proliferation fractions can be trained by abstract computerized images of virtual tissues. Twenty computer‐generated images with randomly distributed blue or red dots were scored by 12 probands (all co‐workers or collaborators of the Institute of Pathology, University of Bonn). Afterward, the probands underwent a training phase during which they received an immediate feedback on the actual rate of positivity after each image. Finally, the initial testing series was rescored. In a second round with 15 different probands, 20 Ki‐67 immunohistochemistry images of tonsil tissue were scored, followed by the same training phase with computer‐generated images, before the immunohistochemistry slides were scored again. Paired t‐tests were used to compare the differences in mean rates pre‐ and post‐training. Concerning computerized images, untrained probands scored the percentages of positive dots with a mean deviation from the true rates of 8.2%. Following training, the same testing series was scored significantly better with a mean deviation of 4.9% (mean improvement 3.3%, p < 0.001). Scoring real immunohistochemistry slides, the training with computerized images also improved correct estimations, albeit to a lesser degree (mean improvement 1%, p = 0.03). Abstract computerized images of virtual tissues may be a useful tool to train and improve the accuracy of pattern recognition involved in semiquantitative scoring of immunohistochemistry slides. As a side results, this study highlights the value of computer‐generated images to verify the performance of image‐analysis software. [ABSTRACT FROM AUTHOR]

• Novel Photonic Device Development : The paper introduces a groundbreaking approach to create a photonic device for chemical applications, utilizing an evanescent wave absorption technique-based optical fiber probe. • Free-base Porphyrin Monolayering: The sensing probe incorporates a Tetraamine porphyrin (TAPP) monolayer, exploiting the unique optical properties of TAPP to enhance chemical sensing capabilities through the optical fiber. • Distinct Optical Patterns: The study focuses on monitoring interactions between TAPP and nitrate and chloride ions in the solution phase, emphasizing on absorption spectra variations of porphyrin molecules significantly in Soret and Q-bands. Distinct optical patterns emerge from the interaction, and the paper thoroughly analyses ion-dependent shifts in both the bands across eight samples of nitrates and chlorides. • Spectral Pattern Recognition: It employs unsupervised learning algorithms for spectral pattern recognition, allowing the discernment of ion groups based on their unique optical signatures. The use of PCA and Dtrend techniques helps to extract significant data to cluster different compounds. The sensor probe however exhibits remarkable sensitivity to nitrate and chloride ions, showcasing an impressive rapid response time of 2-3 s. • Practical Implications: The findings highlight the potential practical applications of the developed photonic device, emphasizing its utility in environmental monitoring and chemical analysis. The study positions porphyrin-based optical fiber sensor as a valuable technology, demonstrating their efficacy in detecting and analysing chemical compounds, particularly nitrate and chloride ions. The paper introduces a novel approach to develop a photonic device for chemical applications using an evanescent wave absorption technique-based optical fiber probe layered with a Tetraamine porphyrin (TAPP) monolayer exploiting the optical properties of TAPP. The interactions between the TAPP and nitrate and chloride ions in the solution phase were monitored and the changes in the absorption spectra of TAPP molecules were recorded. Distinct optical patterns emerge, and ion-dependent shifts in the Soret and Q-bands are analyzed for eight samples of nitrates and chlorides. Principal Component Analysis (PCA) is employed to discern ion groups based on the spectral pattern recognition. These fiber probes exhibit sensitivity to nitrate and chloride ions, with enhanced selectivity, robustness, and a rapid response time of 2–3 seconds. The study demonstrates the emergence of porphyrin-based optical fiber sensors as a useful sensor technology in environmental monitoring. [ABSTRACT FROM AUTHOR]