Real-time Processing Research Articles

Real-time analysis of the biomolecular interaction between gelsolin and Aβ1-42 monomer and its implication for Alzheimer's disease

Biomolecular interaction acts a pivotal part in understanding the mechanisms underlying the development of Alzheimer's disease (AD). Herein, we built a biosensing platform to explore the interaction between gelsolin (GSN) and different β-amyloid protein 1-42 (Aβ1-42) species, including Aβ1-42 monomer (m-Aβ), Aβ1-42 oligomers with both low and high levels of aggregation (LLo-Aβ and HLo-Aβ) via dual polarization interferometry (DPI). Real-time molecular interaction process and kinetic analysis showed that m-Aβ had the strongest affinity and specificity with GSN compared with LLo-Aβ and HLo-Aβ. The impact of GSN on inhibiting aggregation of Aβ1-42 and solubilizing Aβ1-42 aggregates was evaluated by circular dichroism (CD) spectroscopy. The maintenance of random coil structure of m-Aβ and the reversal of β-sheet structure in HLo-Aβ were observed, demonstrating the beneficial effects of GSN on preventing Aβ from aggregation. In addition, the structure of m-Aβ/GSN complex was analyzed in detail by molecular dynamics (MD) simulation and molecular docking. The specific binding sites and crucial intermolecular forces were identified, which are believed to stabilize m-Aβ in its soluble state and to inhibit the fibrilization of Aβ1-42. Combined theoretical simulations and experiment results, we speculate that the success of GSN sequestration mechanism and the balance of GSN levels in cerebrospinal fluid and plasma of AD subjects may contribute to a delay in AD progression. This research not only unveils the molecular basis of the interaction between GSN and Aβ1-42, but also provides clues to understanding the crucial functions of GSN in AD and drives the development of AD drugs and therapeutic approaches.

Multi-Domain Joint Synthetic Aperture Radar Ship Detection Method Integrating Complex Information with Deep Learning

With the flourishing development of deep learning, synthetic aperture radar (SAR) ship detection based on this method has been widely applied across various domains. However, most deep-learning-based detection methods currently only use the amplitude information from SAR images. In fact, phase information and time-frequency features can also play a role in ship detection. Additionally, the background noise and the small size of ships also pose challenges to detection. Finally, satellite-based detection requires the model to be lightweight and capable of real-time processing. To address these difficulties, we propose a multi-domain joint SAR ship detection method that integrates complex information with deep learning. Based on the imaging mechanism of line-by-line scanning, we can first confirm the presence of ships within echo returns in the eigen-subspace domain, which can reduce detection time. Benefiting from the complex information of single-look complex (SLC) SAR images, we transform the echo returns containing ships into the time-frequency domain. In the time-frequency domain, ships exhibit distinctive features that are different from noise, without the limitation of size, which is highly advantageous for detection. Therefore, we constructed a time-frequency SAR image dataset (TFSID) using the images in the time-frequency domain, and utilizing the advantages of this dataset, we combined space-to-depth convolution (SPDConv) and Inception depthwise convolution (InceptionDWConv) to propose Efficient SPD-InceptionDWConv (ESIDConv). Using this module as the core, we proposed a lightweight SAR ship detector (LSDet) based on YOLOv5n. The detector achieves a detection accuracy of 99.5 with only 0.3 M parameters and 1.2 G operations on the dataset. Extensive experiments on different datasets demonstrated the superiority and effectiveness of our proposed method.

Development of a Machine Learning Framework for Real-Time PMI (Post-Mortem Interval) Estimation in Field Forensics

Accurate estimation of the Post-Mortem Interval (PMI) is critical in forensic investigations, aiding in determining the time of death. However, traditional PMI estimation methods, often reliant on physiological observations and environmental factors, face significant limitations in accuracy and efficiency, especially in field conditions. This paper presents the development of a machine learning (ML) framework designed for real-time PMI estimation, integrating multimodal sensor data to address the challenges encountered in field forensics. Our framework utilizes environmental and physiological features, including body temperature, ambient humidity, and biochemical decomposition markers, to predict PMI with high precision. The ML model, trained on historical forensic data, is deployed on a real-time processing platform, enabling rapid analysis and decision-making in resource- constrained environments. The system is optimized for field operations, incorporating low-power hardware and edge computing capabilities to provide forensic investigators with reliable PMI estimates on-site. Through a series of controlled experiments simulating forensic scenarios, our framework demonstrates a significant improvement in PMI accuracy compared to traditional methods, while maintaining low latency for real-time applications. This research highlights the potential of machine learning to revolutionize forensic practices, offering a scalable and adaptive solution for time-sensitive investigations. Here are some relevant keywords for the development of a machine learning framework for real-time PMI (Post- Mortem Interval) estimation in field forensics: Keywords: Field Forensics, Real-Time Machine Learning, Body Decomposition Stages, Machine Learning in Forensic Science, Artificial Intelligence for PMI Analysis, Sensor Data in PMI Estimation, Deep Learning for PMI Estimation, Automated Forensic Analysis, Data Acquisition in Field Forensics.

Based on improved joint detection and tracking of UAV for multi-target detection of livestock

In agriculture, specifically livestock monitoring, drones' ability to track multiple targets is essential for advancing the field. However, limited computing resources and unpredictable drone movements often cause issues like ambiguous video frames, object obstructions, and size deviations. These inconsistencies reduce tracking accuracy, making traditional algorithms inadequate for handling drone footage. This study introduces an enhanced deep learning-based multi-target drone tracker framework that enables real-time processing. The proposed method combines object detection and tracking by leveraging consecutive frame pairs to extract and share features, enhancing computational efficiency. It employs diverse loss functions to address class and sample distribution imbalances and includes a composite deblurring module to enhance detection accuracy. Object association utilizes a dual regress bounding box technique, aiding in object identification verification and predictive motion. Live tracking is achieved by predicting object locations in subsequent frames, enabling real-time tracking. Evaluation against leading benchmarks shows that the system improves precision and speed, achieving a 4.3 % increase in Multi-Object Tracking Accuracy (MOTA) and a 7.7 % boost in F1 score.

Energy-aware bio-inspired spiking reinforcement learning system architecture for real-time autonomous edge applications.

Mobile, low-cost, and energy-aware operation of Artificial Intelligence (AI) computations in smart circuits and autonomous robots will play an important role in the next industrial leap in intelligent automation and assistive devices. Neuromorphic hardware with spiking neural network (SNN) architecture utilizes insights from biological phenomena to offer encouraging solutions. Previous studies have proposed reinforcement learning (RL) models for SNN responses in the rat hippocampus to an environment where rewards depend on the context. The scale of these models matches the scope and capacity of small embedded systems in the framework of Internet-of-Bodies (IoB), autonomous sensor nodes, and other edge applications. Addressing energy-efficient artificial learning problems in such systems enables smart micro-systems with edge intelligence. A novel bio-inspired RL system architecture is presented in this work, leading to significant energy consumption benefits without foregoing real-time autonomous processing and accuracy requirements of the context-dependent task. The hardware architecture successfully models features analogous to synaptic tagging, changes in the exploration schemes, synapse saturation, and spatially localized task-based activation observed in the brain. The design has been synthesized, simulated, and tested on Intel MAX10 Field-Programmable Gate Array (FPGA). The problem-based bio-inspired approach to SNN edge architectural design results in 25X reduction in average power compared to the state-of-the-art for a test with real-time context learning and 30 trials. Furthermore, 940x lower energy consumption is achieved due to improvement in the execution time.

DBDAA: A real-time approach to Dynamic Banker’s Deadlock Avoidance Algorithm with optimized time complexity

Effective resource allocation is crucial in operating systems to prevent deadlocks, especially when resources are limited and non-shareable. Traditional methods like the Banker’s algorithm provide solutions but suffer from limitations such as static process handling, high time complexity, and a lack of real-time adaptability. To address these challenges, we propose the Dynamic Banker’s Deadlock Avoidance Algorithm (DBDAA). The DBDAA introduces real-time processing for safety checks, significantly improving system efficiency and reducing the risk of deadlocks. Unlike conventional methods, the DBDAA dynamically includes processes in safety checks, considerably decreasing the number of comparisons required to determine safe states. This optimization reduces the time complexity to O(n) in the best-case and O(nd) in the average and worst-case scenarios, compared to the O(n2d) complexity of the original Banker’s algorithm. The integration of real-time processing ensures that all processes can immediately engage in safety checks, improving system responsiveness and making the DBDAA suitable for dynamic and time-sensitive applications. Additionally, the DBDAA introduces a primary unsafe sequence mechanism that enhances the acceptability and efficiency of the algorithm by allowing processes to participate in safety checks repeatedly after a predetermined amount of system-defined time. Experimental comparisons with existing algorithms demonstrate the superiority of the DBDAA in terms of reduced safe state prediction time and increased efficiency, making it a robust solution for deadlock avoidance in real-time systems.

Time series forecasting of price of the agricultural products using data science

Purpose. The purpose of our article is to research and forecast prices for agricultural products using the example of potato prices based on the most effective models using data science techniques. Methodology / approach. Various forecasting models are explored, starting from baseline models like decomposition and exponential smoothing models to more advanced techniques such as ARIMA, SARIMA, as well as deep learning models including neural network. The data is split into training and testing sets, and models are validated using cross-validation techniques and optimised through hyperparameter tuning. Model performance is evaluated using metrics such as MAE, MSE, RMSE, and MAPE. The selected model is then used to generate future price forecasts, with uncertainty quantified through confidence intervals. Results. The study successfully applied advanced data science techniques to forecast potato prices, leveraging a range of effective models. By analysing historical price data and using various forecasting methods, the research identified the most accurate models for predicting future price trends. The results demonstrate that the selected models can provide reliable forecasts. In particular, the results showed that the model could achieve good forecast results when applied to real problems and, thus, can be effectively used for forecasting tasks especially considering seasonality. In addition, it should be noted that the model has a higher prediction accuracy at the time intervals closest to the original data. The obtained results support using both models simultaneously for forecasting, which can compensate for the shortcomings of each of them. The models can be used separately, to more accurately predict the values for the required period, or a combination of them is also possible. Originality / scientific novelty. The study’s originality lies in development of methods for effectively accounting for seasonality in agricultural price data, such as using seasonal decomposition techniques or more advanced techniques that combine statistical and data science approaches. The novelty implies the implementation of real-time data processing and forecasting system allows for the timely prediction of price changes, enabling stakeholders to make more informed decisions. Practical value / implications. Forecasting potato prices holds significant practical value for various stakeholders. For farmers, accurate forecasts enable informed decisions on the optimal times to plant, harvest, and sell their crops, thereby optimising their profits. In the supply chain, distributors and retailers can use these forecasts to manage inventory more effectively and plan contracts, reducing waste and avoiding shortages. Policymakers benefit from forecasts by anticipating market fluctuations and stabilising prices, which supports both consumers and producers. For consumers, stable pricing ensures better budgeting and helps avoid sudden price spikes, making essential foods more affordable. Overall, accurate price forecasting enhances market efficiency by reducing uncertainty and aiding investors in managing risk.

Obstacle feature point detection method for real-time processing by monocular camera and line laser

In this study, we propose a method for detecting feature points of obstacles by real-time processing using a monocular camera and a laser. Specifically, we propose an algorithm that detects laser beams emitted on obstacles by binarization, transforms pixel coordinates into distance using triangulation by a laser beam that is obliquely incident on the ground, and estimates the placement angle of the obstacle in real-time processing. No method has been devised to detect the placement angle of obstacles using a monocular camera and a laser. Furthermore, the proposed method does not calculate the linear distance from the camera, but from the position where the camera is moved horizontally parallel to the obstacle in front of the camera, thus detecting the placement angle independently of the placement position of the obstacle. As a result, it was confirmed that the placement angles of obstacles could be detected with a maximum error of 6∘\\documentclass[12pt]{minimal} \\usepackage{amsmath} \\usepackage{wasysym} \\usepackage{amsfonts} \\usepackage{amssymb} \\usepackage{amsbsy} \\usepackage{mathrsfs} \\usepackage{upgreek} \\setlength{\\oddsidemargin}{-69pt} \\begin{document}$$6^{\\circ }$$\\end{document} in six placement positions. Therefore, we succeeded in automatically detecting the placement angle of obstacles in real-time, independent of the illumination of the measurement environment, the reflectance of the obstacle.

Digital Twin Approach for Operation and Maintenance of Transportation System-Systematic Review.

There is a growing need to implement modern technologies, such as digital twinning, to improve the efficiency of transport fleet maintenance processes and maintain company operational capacity at the required level. A comprehensive review of the existing literature is conducted to address this, offering an up-to-date analysis of relevant content in this field. The methodology employed is a systematic literature review using the Primo multi-search tool, adhering to the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines. The selection criteria focused on English studies published between 2012 and 2024, resulting in 201 highly relevant papers. These papers were categorized into seven groups: (a) air transportation, (b) railway transportation, (c) land transportation (road), (d) in-house logistics, (e) water and intermodal transportation, (f) supply chain operation, and (g) other applications. A notable strength of this study is its use of diverse scientific databases facilitated by the multi-search tool. Additionally, a bibliometric analysis was performed, revealing the evolution of DT applications over the past decade and identifying key areas such as predictive maintenance, condition monitoring, and decision-making processes. This study highlights the varied levels of adoption across different transport sectors and underscores promising areas for future development, particularly in underrepresented domains like supply chains and water transport. Additionally, this paper identifies significant research gaps, including integration challenges, real-time data processing, and standardization needs. Future research directions are proposed, focusing on enhancing predictive diagnostics, automating maintenance processes, and optimizing inventory management. This study also outlines a framework for DT in transportation systems, detailing key components and functionalities essential for effective maintenance management. The findings provide a roadmap for future innovations and improvements in DT applications within the transportation industry. This study ends with conclusions and future research directions.

Health Community 4.0: An Innovative Multidisciplinary Solution for Tailored Healthcare Assistance Management.

This paper presents a co-designed solution aimed at improving community healthcare assistance management. To enhance patients' self-awareness of their health status, encouraging active participation in care process, and to support nurses in delivering patient-centered care, we have developed an innovative platform with a highly customized app. This platform was designed using a multi-disciplinary, bottom-up approach. Patient data collection and processing facilitate the automation of care timeline planning, based on real-time patients' needs and the available resources. To achieve this goal, different components have been considered: real-time health data collection and processing, patient care planning, decision support for nurses, secure communication for data transmission, and user-friendly interfaces to ensure easy access to platform functionalities.

An FPGA-Based YOLOv5 Accelerator for Real-Time Industrial Vision Applications.

The You Only Look Once (YOLO) object detection network has garnered widespread adoption in various industries, owing to its superior inference speed and robust detection capabilities. This model has proven invaluable in automating production processes such as material processing, machining, and quality inspection. However, as market competition intensifies, there is a constant demand for higher detection speed and accuracy. Current FPGA accelerators based on 8-bit quantization have struggled to meet these increasingly stringent performance requirements. In response, we present a novel 4-bit quantization-based neural network accelerator for the YOLOv5 model, designed to enhance real-time processing capabilities while maintaining high detection accuracy. To achieve effective model compression, we introduce an optimized quantization scheme that reduces the bit-width of the entire YOLO network-including the first layer-to 4 bits, with only a 1.5% degradation in mean Average Precision (mAP). For the hardware implementation, we propose a unified Digital Signal Processor (DSP) packing scheme, coupled with a novel parity adder tree architecture that accommodates the proposed quantization strategies. This approach efficiently reduces on-chip DSP utilization by 50%, offering a significant improvement in performance and resource efficiency. Experimental results show that the industrial object detection system based on the proposed FPGA accelerator achieves a throughput of 808.6 GOPS and an efficiency of 0.49 GOPS/DSP for YOLOv5s on the ZCU102 board, which is 29% higher than a commercial FPGA accelerator design (Xilinx's Vitis AI).

Linearized wavefront sensing model for aberration retrieval from low-frequency Fourier coefficients

This paper proposes and demonstrates a linearized model for phase diversity wavefront sensing, facilitating real-time processing and much less data required for training. Specifically, we find that the low-frequency Fourier coefficients of point spread function images are linearly proportional to pupil aberration coefficients under certain conditions. Simulation and experimental results show that the model can greatly reduce the processing time to several milliseconds by merely requiring hundreds of training samples while maintaining a comparatively high accuracy with state-of-the-art methods.

Exploring the structure, metabolism, and biochemistry of the neuronal microenvironment label-free using fast simultaneous multimodal optical microscopy

The technologies to examine the neuronal microenvironment label free remain critically underexplored. There is a gap in our knowledge of underlying metabolic, biochemical, and electrophysiological mechanisms behind several neurological processes at a cellular level, which can be traced to the lack of versatile and high-throughput tools to investigate neural networks. In this paper, four label-free contrasts were explored as mechanisms to study neuronal activity, namely, scattering, birefringence, autofluorescence from metabolic cofactors and molecules, and local biochemistry. To overcome challenges of observing neuronal activity spanning three orders of magnitude in space and time, microscopes had to be developed to simultaneously capture these contrasts quickly, with high resolution, and over a large FOV. We developed versatile autofluorescence lifetime, multiharmonic generation, polarization-sensitive interferometry, and Raman imaging in epi-detection (VAMPIRE) microscopy to simultaneously observe multiple facets of neuronal structure and dynamics. The accelerated computational-imaging-driven acquisition speeds, the utilization of a single light source to evoke all contrasts, the simultaneous acquisition that provides an otherwise impossible multimodal dynamic imaging capability, and the real-time processing of the data enable VAMPIRE microscopy as a powerful imaging platform for neurophotonics and beyond.

Research on improved underwater cable image processing technique based on CNN-GAN

In this study, we propose a CNN-GAN-based real-time processing technique for filtering images of underwater cables used in power systems. This addresses the excessive interference impurities that are frequently observed in images captured by remotely operated vehicles (ROVs). The process begins with the input of the original image into the convolutional neural network (CNN). Subsequently, the training outcomes, which serve as input parameters for the generative adversarial network (GAN), facilitate the filtering process. The system also calculates both the structural similarity index measure (SSIM) and peak signal-to-noise ratio (PSNR), performing model updates via backward propagation. This technique utilizes deep learning technologies to achieve rapid, real-time filtering of underwater cable images. The experimental results reveal that the loss function of the CNN reaches 0.16 with an accuracy of 97.5%, while the loss function of the adversarial GAN network approaches 0.05. Compared with traditional methods such as DDN, JORDOR, RESCAN, and PRENet, the proposed CNN-GAN algorithm exhibits superior performance, as evidenced by the higher PSNR and SSIM values. Specifically, for clear water images, the PSNR reaches 29.86 dB and the SSIM is 0.9045. For severely polluted images, the PSNR is 28.67 dB and the SSIM is 0.8965, while for unevenly illuminated images, the PSNR and SSIM values are 24.37 dB and 0.88, respectively. These enhancements significantly benefit the monitoring and maintenance of power systems.

Leading-Edge Sentiment Analysis: A Survey of Application Context, Challenges and Advanced Techniques

Background: Data is rapidly expanding in today's digital age. The reason for the expansion of data is due i to social media sites. The Internet produces an enormous quantity of unstructured data every second. Numerous users have many opinions and reviews to impart on everything from items and services to common pastimes. Opinions, feelings, attitudes, impressions, etc., concerning subjects, products, and services are collected and analyzed through a method called sentiment analysis. Web-based networking mediums that rely on textual communication can be overwhelming. Understanding human psychology requires the real-time processing of data using techniques like sentiment analysis. Aim: This study provides a thorough examination of the differences between methods of sentiment analysis, as well as its obstacles and emerging trends. The paper exemplifies the analysis's practical uses, examines its challenges, and outlines common methods of conducting it. Objective: The objective of the current overview is to better understand the market, gauge public opinion, and make strategic decisions. In addition, enterprises, governments, and scholars can all benefit from conducting a sentiment analysis. Methods: In this research, we review and categorize the most widely applied methods of deep learning and machine learning for analyzing sentiment. From the paper, we learn that which sentiment analysis technique is the best depends on the data at hand. When confronted with large amounts of data and a lengthy procedure, traditional machine learning-based algorithms flop. The ability to train deep learning models to learn more features using larger datasets is why they currently beat machine learning methodologies. Considerations include textual and temporal context, as well as data volume. Results: Regardless of the fact that the English language has traditionally been the focus of research in this field, other spoken languages have recently attracted a growing amount of interest. The lack of resources for these languages continues to present numerous obstacles. Consequently, it can be an intriguing line of future effort to tackle other natural languages outside English by generating beneficial resources like building databases and addressing the problems with Natural language processing that have been stated in the context of sentiment examination. Conclusion: The difficulties of sentiment analysis are examined as well, with the goal of illuminating potential solutions.

A Channel-Sensing-Based Multipath Multihop Cooperative Transmission Mechanism for UE Aggregation in Asymmetric IoE Scenarios

With the continuous progress and development of technology, the Internet of Everything (IoE) is gradually becoming a research hotspot. More companies and research institutes are focusing on the connectivity and transmission between multiple devices in asymmetric networks, such as V2X, Industrial Internet of Things (IIoT), environmental monitoring, disaster management, agriculture, and so on. The number of devices and business volume of these applications have rapidly increased in recent years, which will lead to a large load of terminals and affect the transmission efficiency of IoE data transmission. To deal with this issue, it has been proposed to perform data transmission via multipath cooperative transmission with multihop transmission. This approach aims to improve transmission latency, energy consumption, reliability, and throughput. This paper designs a channel-sensing-based cooperative transmission mechanism (CSCTM) with hybrid automatic repeat request (HARQ) for user equipment (UE) aggregation mechanism in future asymmetric IoE scenarios, which ensures that IoE devices data can be transmitted quickly and reliably, and supports real-time data processing and analysis. The main contents of this proposed method include strategies of cooperative transmission and redundancy version (RV) determination, a joint combination of decoding process at the receiving side, and a design of transmission priority through ascending offset sort (AOS) algorithm based on channel sensing. In addition, multihop technology is designed for the multipath cooperative transmission strategy, which enables cooperative nodes (CN) to help UE to transmit data. As a result, it can be obtained that CSCTM provides significant advancements in latency and energy consumption for the whole system. It demonstrates improvements in enhanced coverage, improved reliability, and minimized latency.

Real-Time Semantic Segmentation Algorithm for Street Scenes Based on Attention Mechanism and Feature Fusion

In computer vision, the task of semantic segmentation is crucial for applications such as autonomous driving and intelligent surveillance. However, achieving a balance between real-time performance and segmentation accuracy remains a significant challenge. Although Fast-SCNN is favored for its efficiency and low computational complexity, it still faces difficulties when handling complex street scene images. To address this issue, this paper presents an improved Fast-SCNN, aiming to enhance the accuracy and efficiency of semantic segmentation by incorporating a novel attention mechanism and an enhanced feature extraction module. Firstly, the integrated SimAM (Simple, Parameter-Free Attention Module) increases the network’s sensitivity to critical regions of the image and effectively adjusts the feature space weights across channels. Additionally, the refined pyramid pooling module in the global feature extraction module captures a broader range of contextual information through refined pooling levels. During the feature fusion stage, the introduction of an enhanced DAB (Depthwise Asymmetric Bottleneck) block and SE (Squeeze-and-Excitation) attention optimizes the network’s ability to process multi-scale information. Furthermore, the classifier module is extended by incorporating deeper convolutions and more complex convolutional structures, leading to a further improvement in model performance. These enhancements significantly improve the model’s ability to capture details and overall segmentation performance. Experimental results demonstrate that the proposed method excels in processing complex street scene images, achieving a mean Intersection over Union (mIoU) of 71.7% and 69.4% on the Cityscapes and CamVid datasets, respectively, while maintaining inference speeds of 81.4 fps and 113.6 fps. These results indicate that the proposed model effectively improves segmentation quality in complex street scenes while ensuring real-time processing capabilities.

Leveraging Oracle HCM for Enhanced Employee Engagement

As the current corporate environment continues to evolve, employee engagement has emerged as a critical component in determining the success of an organisation. Oracle Human Capital Management (HCM) systems provide a revolutionary chance to improve employee engagement by providing a full suite of tools that are geared to handle different areas of workforce management. This opportunity presents itself as a result of the availability of these systems. In order to provide a more unified and responsive approach to employee engagement, Oracle Human Capital Management (HCM) offers an integrated platform that incorporates basic HR tasks, talent management, workforce management, and analytics. In this article, we investigate how the capabilities of Oracle Human Capital Management (HCM) may be strategically used to cultivate a staff that is more engaged and motivated. The cloud-based design of the system guarantees accessibility and real-time data processing, both of which are crucial for interactions with workers that are timely and relevant. A variety of features that may be adapted to assist engagement activities are provided by the modules that make up Oracle Human Capital Management (HCM). These modules include Employee Central, Talent Management, and Learning Management. For example, the Talent Management module helps to connect employee goals with organisational objectives, which in turn improves both work happiness and performance. While this is going on, the Learning Management module offers individualised training and development possibilities, according to the workers' requirements for professional advancement and their goals for their careers.

Will Transformers change gastrointestinal endoscopic image analysis? A comparative analysis between CNNs and Transformers, in terms of performance, robustness and generalization

Gastrointestinal endoscopic image analysis presents significant challenges, such as considerable variations in quality due to the challenging in-body imaging environment, the often-subtle nature of abnormalities with low interobserver agreement, and the need for real-time processing. These challenges pose strong requirements on the performance, generalization, robustness and complexity of deep learning-based techniques in such safety–critical applications. While Convolutional Neural Networks (CNNs) have been the go-to architecture for endoscopic image analysis, recent successes of the Transformer architecture in computer vision raise the possibility to update this conclusion. To this end, we evaluate and compare clinically relevant performance, generalization and robustness of state-of-the-art CNNs and Transformers for neoplasia detection in Barrett’s esophagus. We have trained and validated several top-performing CNNs and Transformers on a total of 10,208 images (2,079 patients), and tested on a total of 7,118 images (998 patients) across multiple test sets, including a high-quality test set, two internal and two external generalization test sets, and a robustness test set. Furthermore, to expand the scope of the study, we have conducted the performance and robustness comparisons for colonic polyp segmentation (Kvasir-SEG) and angiodysplasia detection (Giana). The results obtained for featured models across a wide range of training set sizes demonstrate that Transformers achieve comparable performance as CNNs on various applications, show comparable or slightly improved generalization capabilities and offer equally strong resilience and robustness against common image corruptions and perturbations. These findings confirm the viability of the Transformer architecture, particularly suited to the dynamic nature of endoscopic video analysis, characterized by fluctuating image quality, appearance and equipment configurations in transition from hospital to hospital. The code is made publicly available at: https://github.com/BONS-AI-VCA-AMC/Endoscopy-CNNs-vs-Transformers.

Development of Digital Image Processing Algorithms via FPGA Implementation

Real-time image processing is one of the fundamental elements in achieving IR 4.0. The rapid development of digital image processing techniques has enabled various applications in fields such as healthcare, transportation, and manufacturing. People are seeking higher-performance image processing as traditional image processing is no longer fulfilling the demands. FPGA-based digital image processing has become one of the choices for the public due to its parallel pipelining, which enables shorter processing time and better performance. Several digital image processing algorithms have been developed in this project, which are gray level transformation, brightness manipulation, contrast adjustment, thresholding, and inversion. They are the most popular algorithms used in digital image processing. Microsoft Paint is used to convert the format of the color input image to bitmap format, followed by MATLAB to convert it into a hexadecimal file to be read and written in FPGA. Platforms such as ModelSim Altera and Intel Quartus II are used to write Verilog HDL for digital image processing algorithms. As a result, five hexadecimal files are obtained from the simulation. The output hexadecimal files are further processed in MATLAB to generate respective images.