Real-time Processing Research Articles

Benchmarking neural radiance fields for autonomous robots: An overview

Neural Radiance Field (NeRF) has emerged as a powerful paradigm for scene representation, offering high-fidelity renderings and reconstructions from a set of sparse and unstructured sensor data. In the context of autonomous robotics, where perception and understanding of the environment are pivotal, NeRF holds immense promise for improving performance. However, few survey has discussed such a potential. To fill this gap, we have collected over 200 papers since the publication of original NeRF in 2020 and present a thorough analysis of how NeRF can be used to enhance the capabilities of autonomous robots. We especially focus on the perception, localization and navigation, and decision-making modules of autonomous robots and delve into tasks crucial for autonomous operation, including 3-dimensional reconstruction, segmentation, pose estimation, simultaneous localization and mapping, navigation and planning, and interaction. Our survey meticulously benchmarks existing NeRF-based methods, comparing their reported performance, and providing insights into their strengths and limitations. Moreover, we target the existing challenges of applying NeRF in autonomous robots, including real-time processing, sparse input views, and explore promising avenues for future research and development in this domain. We especially discuss potential of integrating advanced deep learning techniques like 3-dimensional Gaussian splatting, large language models, and generative artificial intelligence. This survey serves as a roadmap for researchers seeking to leverage NeRF to empower autonomous robots, paving the way for innovative solutions that can navigate and interact seamlessly in complex environments.

A lightweight approach to real-time speaker diarization: from audio toward audio-visual data streams

This manuscript deals with the task of real-time speaker diarization (SD) for stream-wise data processing. Therefore, in contrast to most of the existing papers, it considers not only the accuracy but also the computational demands of individual investigated methods. We first propose a new lightweight scheme allowing us to perform speaker diarization of streamed audio data. Our approach utilizes a modified residual network with squeeze-and-excitation blocks (SE-ResNet-34) to extract speaker embeddings in an optimized way using cached buffers. These embeddings are subsequently used for voice activity detection (VAD) and block-online k-means clustering with a look-ahead mechanism. The described scheme yields results similar to the reference offline system while operating solely on a CPU with a low real-time factor (RTF) below 0.1 and a constant latency of around 5.5 s. In the next part of the work, our research moves toward much more demanding and complex real-time processing of audio-visual data streams. For this purpose, we extend the above-mentioned scheme for audio data processing by adding an audio-video module. This module utilizes SyncNet combined with visual embeddings for identity tracking. Our resulting multi-modal SD framework then combines the outputs from audio and audio-video modules by using a new overlap-based fusion strategy. It yields diarization error rates that are competitive with the existing state-of-the-art offline audio-visual methods while allowing us to process various audio-video streams, e.g., from Internet or TV broadcasts, in real-time using GPU and with the same latency as for audio data processing.

An IoT-Enabled System for Prioritizing Emergency Vehicles and Detecting Accidents in Smart Traffic Management

Abstract—With the growth of urban areas and the rising number of vehicles, managing traffic congestion has become increasingly difficult, especially when it comes to ensuring prompt clearance for emergency vehicles like ambulances, fire trucks, and police cars. This paper introduces a Smart Traffic Management System powered by IoT, which leverages camera modules and machine learning algorithms to identify and prioritize emergency vehicles. Ambulances receive the highest priority, followed by fire trucks and police vehicles. The system adapts traffic signals dynamically by detecting emergency vehicles in real-time, and also enhances traffic flow using IR sensors to monitor vehicle density. Additionally, the system includes an Accident Detection Module that leverages intelligent accelerometer sensors to detect road accidents, using data on speed, pressure, and distance. This holistic approach ensures that emergency lanes are cleared efficiently and improves traffic mobility in urban environments. The proposed system significantly enhances the efficiency of traffic management compared to conventional methods, offering a comprehensive solution for reducing congestion, improving emergency response times, and minimizing accident fatalities. Index Terms—IoT, smart traffic management, machine learning, emergency vehicle detection, traffic signal optimization, IR sensors, accident detection, accelerometer sensors, real-time data processing, urban mobility, SCATS (Sydney Coordinated Adaptive Traffic System), IoT-Enabled Smart Traffic Management System (ISTMS).

Cloud-Native Microservices for Real-Time Data Systems: A Technical Deep Dive

Cloud-native microservices have emerged as a transformative solution for organizations facing the challenges of real-time data processing and analysis in today's digital landscape. This comprehensive article explores the fundamental components of cloud-native architectures, from their foundational principles to implementation best practices. The article examines how containerization, orchestration, and event-driven architectures enable unprecedented scalability and efficiency in handling modern data processing requirements. Through a detailed examination of microservices as building blocks, the research demonstrates how these architectural patterns significantly improve system reliability, performance, and maintainability. The article encompasses advanced monitoring and observability practices, highlighting their crucial role in maintaining robust cloud-native systems. By investigating real-world implementations across various industries, this study provides valuable insights into the practical benefits and challenges of adopting cloud-native approaches.

Siamese neural network improves the performance of a convolutional neural network in colloidal self-assembly state classification.

Identifying the state of the colloidal self-assembly process is critical to monitoring and controlling the system into desired configurations. Recent application of convolutional neural networks with unsupervised clustering has shown a comparable performance to conventional approaches, in representing and classifying the states of a simulated 2D colloidal batch assembly system. Despite the early success, capturing the subtle differences among similar configurations still presents a challenge. To address this issue, we leverage a Siamese neural network to improve the accuracy of the state classification. Results from a Brownian dynamics-simulated electric field-mediated colloidal self-assembly system and a magnetic field-mediated colloidal self-assembly system demonstrate significant improvement from the original convolutional neural network-based approach. We anticipate the proposed improvement to further pave the way for automated monitoring and control of colloidal self-assembly processes in real time and real space.

Gesture Based Control Technology for Drones: Challenges and Solutions

Gesture-based control systems have become an innovative approach in human-computer interaction, enabling users to operate devices like UAVs using hand movements. These systems generally consist of three main components: sensors to capture movements, gesture recognition algorithms to interpret them, and control interfaces to execute actions. However, several challenges still exist that limit the performance and practicality of these systems. Key problems include accuracy in varying environmental conditions, latency issues caused by sensor overload, and user ergonomics, which can affect usability over time. For instance, sensors like optical cameras and IMUs may struggle in low-light or complex environments, leading to errors in gesture interpretation. Moreover, real-time processing is essential for seamless control, but the high computational demands of machine learning algorithms often lead to delays. This paper explores these challenges in detail and presents solutions, such as using multi-sensor setups to improve data accuracy, implementing adaptive algorithms that can learn and adjust to user behavior, and refining control interfaces with real-time feedback mechanisms to enhance the user experience. Addressing these issues will make gesture-based control systems more efficient, reliable, and user-friendly, enabling wider application across industries like virtual reality, robotics, and military operations.

Enhancing Real-Time Vision Systems: Integrating Dynamic Vision Sensors with Vision Transformers to Increase Computational Efficiency

Abstract. Optimizing and understanding vision systems' computational accuracy is crucial as they become increasingly integrated into daily life. Dynamic Vision Sensors (DVS) and Vision Transformers (ViTs) lead computer vision technology with efficient object recognition and image processing. However, DVS data's high computational complexity poses a problem for its real-time implementations. Merging these technologies can enhance vision system performance in dynamic environments and optimize real-time DVS processing. In our work, we use a ViT architecture to classify the DVS 128 dataset and compare our results with existing works using SNNs. We analyze how our method affects accuracy and loss, experimenting with different DVS-to-ViT input patch sizes. Our results show that the large patch size of 32x32 pixels has better accuracy and smaller loss than the 4x4 pixel patches as epochs increase. Our method also achieved a high 98.4% accuracy and low 0.22 loss within five epochs, significantly outperforming previous works averaging 93.13% accuracy over more epochs. These results highlight ViTs' large potential for real-time DVS data classification in applications that require high accuracy, like autonomous vehicles and surveillance systems.

Data Processing of Used Car Export Service Flow Based on Zookeeper + Kafka

The used car export industry has shown a trend of rapid growth in recent years. The export of used cars involves data pertaining to vehicle information, market demand, logistics, customs, and numerous other aspects. These data are constantly generated in the form of streams, posing challenges to the data processing capabilities of service platforms. Traditional data processing methods struggle to meet the requirements of real-time and large-scale data processing, thus necessitating an efficient streaming data processing architecture. Based on this, this paper proposes a streaming data processing scheme for a used car export service platform utilizing Zookeeper and Kafka. It describes the architecture design and data processing flow, analyzes the advantages of the scheme in terms of reliability, scalability, and real-time performance, and demonstrates its effectiveness through experiments and case studies.

Exploring the Effects of Attention Mechanisms on Road Crack Detection and Classification

Pavement cracks are a prevalent form of roadway distress that pose significant safety hazards, necessitating prompt detection and repair. Due to the extensive road network, traditional image processing-based crack detection methods exhibit limitations in recognition accuracy and speed. To address these challenges, this paper proposes an efficient pavement crack detection model based on YOLOv8, termed YOLOv8-ACD (YOLOv8 - Attention for Cracks Detecting), which integrates a global attention mechanism. YOLOv8-ACD enhances detection efficiency and accuracy by focusing on crack identification while filtering out most irrelevant information. We evaluated YOLOv8-ACD on the RDD2022 dataset, and experimental results demonstrate significant improvements in key performance metrics, such as F1-Score and mean Average Precision (mAP), compared to the original YOLOv8 and other mainstream models. The real-time processing capability of this model makes it suitable for practical road inspection and maintenance, effectively reducing the workload of maintenance personnel and enhancing roadway safety.

Progress and Application of Unsupervised Feature Extraction Methods

Unsupervised feature extraction is crucial in machine learning and data mining for handling high-dimensional and unlabeled data. However, existing methods often ignore feature relationaships, resulting in suboptimal feature subsets. This paper reviews the current state of unsupervised feature extraction methods, discussing the limitations of traditional methods such as Principal Component Analysis (PCA) and Independent Component Analysis (ICA), particularly in terms of interpretability, sensitivity to outliers, and computational resource challenges. In recent years, improvement strategies such as information theory, sparse learning, and deep learning (e.g., deep autoencoders and generative adversarial networks) have significantly progressed in feature extraction. This paper analyzes the practical applications of these methods in image processing, gene analysis, text mining, and network security. For example, in image processing, deep autoencoder-based methods such as Matrix Capsules with EM Routing can effectively extract key features from complex images. In text mining, unsupervised feature selection methods combined with generative adversarial networks significantly improve the efficiency of processing high-dimensional text data. Additionally, this paper explores future research directions such as multimodal data processing, improving real-time processing capabilities, and integration with other machine learning techniques (e.g., reinforcement learning, transfer learning), providing insights for further development of unsupervised feature extraction technologies.

Optimizing extrusion processes and understanding conformational changes in itraconazole amorphous solid dispersions using in-line UV–Vis spectroscopy and QbD principles

This paper presents a comprehensive investigation of the manufacturing of itraconazole (ITZ) amorphous solid dispersions (ASDs) with Kolllidon® VA64 (KVA64) using hot-melt extrusion (HME) and in-line process monitoring, employing a Quality by Design (QbD) approach. A sequential Design of Experiments (DoE) strategy was utilized to optimize the manufacturing process, with in-line UV–Vis spectroscopy providing real-time monitoring. The first DoE used a fractional factorial screening design to evaluate critical process parameters (CPPs), revealing that ITZ concentration had the most significant impact on the product quality attributes. The second DoE, employing a central composite design, explored the interactions between feed rate and screw speed, using torque and absorbance at 370 nm as responses to develop a design space. Validation studies confirmed process robustness across multiple days, with stable in-line UV–Vis spectra and consistent product quality using 30 % ITZ, 300 rpm, 150 °C and 7 g/min as the optimized process conditions. Theoretical and experimental analyses indicated that shifts in UV–Vis spectra at different ITZ concentrations were due to conformational changes in ITZ, which were confirmed through density functional theory (DFT) calculations and infrared spectroscopy. This work offers novel insights into the production and monitoring of ITZ-KVA64-ASDs, demonstrating that in-line UV–Vis spectroscopy is a powerful tool for real-time process monitoring and/or control.

Advanced Deep Learning Approach for Smart Home Appliance Identification Using Recurrent Neural Networks with LSTM

In the Internet of Things (IoT) domain, vast numbers of smart devices are interconnected, generating large volumes of data requiring advanced management mechanisms. One major challenge in smart environments is the ability to accurately distinguish and categorize the various types of objects within these systems. To address this issue, the study introduces a recurrent neural network (RNN) model designed for classifying data from smart home devices. Using a dataset from Kaggle, the research outlines the processes of data collection, loading, normalization, and model development. The RNN, enhanced with long short-term memory (LSTM) layers, was trained and evaluated, showing notable improvements in training and validation accuracy over ten epochs. The model achieved a test accuracy of 83.25%, a loss of 35.4%, a precision of 85%, and a recall of 81%. The evaluation of the model on the test set includes a detailed analysis using ROC curves, area under the curve (AUC) scores for multi-class classification, and a confusion matrix. With an AUC score of 0.9896, the model demonstrated exceptional performance in accurately classifying IoT device categories. These results suggest that the LSTM-equipped RNN offers strong learning efficiency and generalization, making it a highly suitable approach for IoT device classification. Additionally, the article explores the concept of IoT and reviews recent advancements in using deep learning models across various IoT sectors, including smart homes, industrial systems, and healthcare. Future research could aim to improve the model’s real-time processing abilities and scalability and incorporate a wider variety of IoT data types to enhance its practical applications and expand its utility across more IoT environments.

Visual insights into translation: demystifying trends of adopting eye-tracking techniques in translation studies.

The increasing use of eye-tracking techniques in translation studies offers valuable insights into cognitive processes and behavioral strategies of translators, reflecting a significant trend within cognitive linguistics and translator training methodologies. This review harnesses quantitative bibliometric analysis through Bibliometrix R-package with qualitative content assessment to evaluate the trajectory and thematic evolution of eye-tracking research in translation studies. Through a dataset from the Web of Science, 56 articles were analyzed, revealing distinct thematic dimensions and trend dynamics. The analysis revealed that eye-tracking is increasingly pivotal in exploring the cognitive and technological dimensions of translation. Central themes include interactions with translation tools, machine translation, and human-computer interaction, highlighting the importance of cognitive research in technology-driven translation. Niche areas such as English-Chinese translation and online consultation suggest specialized topics that warrant further investigation. Additionally, emerging themes like cognitive load and sight translation demonstrate a shift toward exploring real-time translation processing. Declining traditional topics, such as broader translation theories, indicate a growing integration of cognitive research with technological advancements. These findings elucidate the growth and diversification of eye-tracking applications in translation studies, emphasizing the method's importance in both academic research and practical applications, thereby informing future studies and enhancing translator training programs.

Real-Time Milk Quality Control Using Multi-Spectral Sensing and Edge Computing: Advancing On-Site Detection of Milk Components with XGBoost

This study explores the use of edge computing technologies to enhance the quality control processes in the dairy industry. Traditional milk quality control methods can be time-consuming and sometimes inadequate, whereas this new approach offers real-time data processing and rapid decision-making capabilities. The objective of the study is to assess the effectiveness of evaluating various spectral characteristics of milk in predicting critical parameters such as protein and fat content. In this research, a multi-channel sensor capable of collecting spectral data at various wavelengths was utilized. The collected data were processed using advanced machine learning models, where XGBoost and other regression models were assessed for their accuracy in predicting protein and fat content. The findings demonstrate the suitability of this technology for quality control in the dairy industry. The results reveal that edge computing-based systems can determine milk quality more quickly and accurately. This technology holds significant potential for overcoming the challenges faced in milk quality control, particularly in developing countries. This study provides valuable insights into how the use of edge computing can enhance operational efficiency and ensure product quality in the dairy industry. This research represents an important step towards developing more effective quality control mechanisms in the dairy industry and aims to establish a robust foundation for future studies. Recommendations focus on the adaptation of this technology to other food safety applications and its diversification for widespread industrial use.

Leveraging SAP BW4HANA for Scalable Data Warehousing in Large Enterprises

In the era of big data, large enterprises face the challenge of efficiently managing and analyzing vast volumes of information to derive actionable insights. SAP BW/4HANA emerges as a powerful solution that facilitates scalable data warehousing, enabling organizations to harness their data for strategic decision-making. This paper explores the key features and advantages of SAP BW/4HANA, highlighting its capability to integrate real-time data processing, advanced analytics, and enhanced data modeling. By leveraging the in-memory computing power of HANA, enterprises can significantly reduce data processing times and improve query performance, thus enabling faster insights and informed decision-making. The paper further discusses the architectural enhancements that BW/4HANA offers, such as simplified data flows, streamlined data integration, and the use of modern user interfaces, which collectively contribute to a more agile data warehousing environment. Moreover, the flexibility of deploying BW/4HANA on-premise or in the cloud allows organizations to scale their data warehousing capabilities in alignment with their growth strategies. Through case studies and real-world examples, this research underscores the effectiveness of SAP BW/4HANA in addressing the complexities of data warehousing in large enterprises, emphasizing its role in fostering a data-driven culture. Ultimately, this paper posits that leveraging SAP BW/4HANA can significantly enhance operational efficiency and support organizations in achieving their strategic objectives in an increasingly data-centric business landscape.

Integrating Robotics in Bioprinting: Advancements in 3D-Printed Tissue and Organ Engineering

The integration of robotics in bioprinting is revolutionizing the field of tissue and organ engineering, enabling unprecedented precision, scalability, and complexity in 3D-printed biological structures. This research explores the advancements brought about by robotic systems in bioprinting processes, focusing on their role in enhancing the fabrication of tissues and organs with intricate architectures and functional properties. Key areas of investigation include robotic-assisted multi-material deposition, real-time process monitoring, and adaptive printing techniques that ensure high fidelity and cell viability. The study also examines the incorporation of robotics into scalable bioprinting workflows for large-scale tissue engineering and transplantable organ production. Ethical considerations, such as regulatory challenges and equitable access, are addressed to highlight the societal implications of these innovations. By bridging robotics, bioengineering, and regenerative medicine, this research underscores the transformative potential of robotic-assisted bioprinting in addressing critical healthcare challenges, including organ shortages and personalized medicine. With advancements in precision and adaptability, robotic systems are poised to reshape the future of bioprinting, paving the way for breakthroughs in tissue engineering and regenerative therapies.

Data-Driven autonomous printing process optimization and real-time abnormality identification in aerosol jet-deposited droplet morphology

ABSTRACT Aerosol Jet Printing (AJP) is a digital direct ink writing technology, which excels in maskless patterning and fine conductive line deposition. However, its potential in droplet-based printing remains largely unexplored, which presents a unique opportunity to pioneer advances in sectors that require precise droplet control. In this research, a novel data-driven approach integrating representative deep learning and machine learning technologies is developed to optimise droplet deposition in AJP. In the proposed method, a stepwise machine learning approach is applied to refine and model droplet morphology in AJP, ensuring systematic process optimisation before deposition. A convolutional neural network (CNN) model is then deployed for real-time process monitoring based on droplet morphology, which facilitates the detection of droplet anomalies during printing. In the subsequent experiments, the autonomous optimisation of process variables and abnormality identification achieved accuracies of 96.1% and 95.5%, respectively, highlighting its potential for droplet deposition optimisation in the AJP process.

Webcams as Windows to the Mind? A Direct Comparison Between In-Lab and Web-Based Eye-Tracking Methods.

There is a growing interest in the use of webcams to conduct eye-tracking experiments over the internet. We assessed the performance of two webcam-based eye-tracking techniques for behavioral research: manual annotation of webcam videos (manual eye-tracking) and the automated WebGazer eye-tracking algorithm. We compared these methods to a traditional infrared eye-tracker and assessed their performance in both lab and web-based settings. In both lab and web experiments, participants completed the same battery of five tasks, selected to trigger effects of various sizes: two visual fixation tasks and three visual world tasks testing real-time (psycholinguistic) processing effects. In the lab experiment, we simultaneously collected infrared eye-tracking, manual eye-tracking, and WebGazer data; in the web experiment, we simultaneously collected manual eye-tracking and WebGazer data. We found that the two webcam-based methods are suited to capture different types of eye-movement patterns. Manual eye-tracking, similar to infrared eye-tracking, detected both large and small effects. WebGazer, however, showed less accuracy in detecting short, subtle effects. There was no notable effect of setting for either method. We discuss the trade-offs researchers face when choosing eye-tracking methods and offer advice for conducting eye-tracking experiments over the internet.

Accelerating Molecular Dynamics Simulations Using Socket-Based Interprocess Communication.

Molecular dynamics (MD) simulations are essential for studying the time evolution of molecular systems. Still, their efficiency is often bottlenecked by file-based interprocess communication (IPC) between MD and electronic structure programs. We present a socket-based IPC implementation that dramatically accelerates MD simulations, reducing the computational time by >10-fold compared to those of traditional file-based methods. Our approach, applied to nonadiabatic molecular dynamics with the Newton-X program, eliminates disk read/write overhead, allowing for faster simulations over longer time scales. This method opens the door to more efficient high-throughput simulations, providing new opportunities for exploring complex molecular processes in real time.

A Roadmap to Systematic Review: Evaluating the Role of Data Networks and Application Programming Interfaces in Enhancing Operational Efficiency in Small and Medium Enterprises

The adoption of Data Networks and Application Programming Interfaces (APIs) has become crucial for small and medium enterprises (SMEs) to streamline operations, improve efficiency, and reduce costs. However, SMEs often face challenges such as resource limitations and security vulnerabilities, which hinder their ability to fully leverage these technologies. This systematic review examines the role of Data Networks and APIs in enhancing operational efficiency within SMEs, focusing on key metrics such as speed, cost reduction, scalability, and security challenges. Following PRISMA 2020 guidelines, we conducted a systematic search across multiple databases including Web of Science, Scopus, IEEE Xplore, and Google Scholar. Studies published between 2014 and 2024, focused on SMEs, and addressing the role of Data Networks and APIs in operational efficiency were included. A total of 49 studies met the inclusion criteria and were analyzed for key outcomes related to operational efficiency, cost-effectiveness, and security risks. The review found that Data Networks and APIs significantly improve operational efficiency by increasing process speed (12% increase), reducing operational costs (8% reduction), and enhancing overall productivity. However, security challenges, particularly related to API vulnerabilities, were a major concern, with cyberattacks on APIs increasing by 400% in Q1 2023 alone. Despite these risks, the benefits of implementing Data Networks and APIs in SMEs, particularly in terms of scalability and real-time data processing, were evident across industries. Data Networks and APIs offer substantial improvements in operational efficiency for SMEs, although security remains a significant challenge. Future efforts should focus on developing security frameworks tailored to SMEs while maintaining the operational benefits of these technologies. Further research is needed to explore scalable and secure API models for SMEs.