Redundant Processing Research Articles

A lightweight Transformer-based visual question answering network with Weight-Sharing Hybrid Attention

Recent advances show that Transformer-based models and object detection-based models play an indispensable role in VQA. However, object detection-based models have significant limitations due to their redundant and complex detection box generation process. In contrast, Visual and Language Pre-training (VLP) models can achieve better performance, but require high computing power. To this end, we present Weight-Sharing Hybrid Attention Network (WHAN), a lightweight Transformer-based VQA model. In WHAN, we replace the object detection network with Transformer encoder and use LoRA to solve the problem that the language model cannot adapt to interrogative sentences. We propose Weight-Sharing Hybrid Attention (WHA) module with parallel residual adapters, which can significantly reduce the trainable parameters of the model and we design DWA and BVA modules that can allow the model to perform attention operations from different scales. Experiments on VQA-v2, COCO-QA, GQA, and CLEVR datasets show that WHAN achieves competitive performance with far fewer trainable parameters.

Packet reception algorithm for redundant data links in transport drones

AbstractIn a drone system with dual data links, this article presents a redundant data processing algorithm that can minimize flight control instability without increasing the weight and price of the aircraft by software processing of duplicate received messages.

[formula omitted] : An adaptive data aggregation algorithm for IoT infrastructure

In IoT infrastructure, high-frequency sensing and subsequent transmission of sensed data to computational facilities can lead to redundant data storage and processing, consuming significant storage and processing capacity. As a result, the IoT infrastructure needs more data transmission cycles, leading to data redundancy and low network up-time due to the drainage of limited battery capacity. Conversely, if the data is communicated at a lower rate, it may cause absolute data delivery to the processing unit, which is useless. As a result, a well-designed data aggregation algorithm is required. This paper proposes the ADA2−IoT, an Adaptive Data Aggregation Algorithm for IoT Infrastructure tailored to optimize parameters such as low data redundancy, limited data communication cycles, and high IoT infrastructure up times. The proposed algorithm consists of two key components: the Route Data Aggregator (RDA) performs aggregation during data transit towards the Edge node or gateway, and the Node Data Aggregator (NDA) performs data aggregation during capturing or sensing data. The algorithm employs metrics like Age of Information (AoI) and data freshness factor during the node and route data aggregation phase to capture and timely deliver data to the Edge node, where this data is processed for informed decision-making. The proposed algorithm was efficiently tested on a simulation and IoT hardware deployment environment. Both simulation and hardware results demonstrate a substantial improvement in QoS parameters, such as a decrease in data redundancy and packet exchanges, leading to considerable energy savings and prolonging the lifespan of IoT infrastructure.

A Hybrid Dual Stream ProbSparse Self-Attention Network for spatial–temporal photovoltaic power forecasting

Growing energy demand and increasing environmental challenges underscore the importance of precise forecasts for photovoltaic (PV) operations in renewable energy generation systems. At this stage, it is mainstream to combine both temporal and spatial factors to forecast PV power generation. However, there are fewer studies that consider factors at very large spatial scales. This paper proposes Hybrid Dual Stream ProbSparse Self-Attention Network (HDSPAN), a novel spatial–temporal photovoltaic power forecasting network architecture that can solve the above limitations. The model implements an encoder–decoder approach that extracts the required spatial–temporal information through a dual stream distilling mechanism. In addition, the ProbSparse self-attention mechanism is employed to improve model efficiency and reduce repetitive and redundant information processing. The hyperparameters are optimized using Tree-structured Patzen estimator to improve forecasting outcomes. This paper demonstrates the effectiveness of spatial–temporal PV forecasting by using ERA5 reanalyzed PV data as a case study. Our results show that the HDSPAN model achieves a 10% higher forecasting accuracy compared to the baseline models, significantly advancing PV power forecasting.

Characteristics of Neutrophil Migration and Function in Acute Inflammation Induced by Zymosan and Carrageenan in the Mice Air Pouch Model.

Deciphering the complex and redundant process of acute inflammation remains challenging. The failure of numerous clinical trials assessing anti-inflammation agents which had promising preclinical effects inevitably questions the validity of current animal models of inflammation. This study aimed to better understand the process of immune inflammatory response and to select more suitable models to evaluate the effect of potential anti-inflammatory drugs. Zymosan and λ-carrageenan are the most used representatives of particulate and soluble irritants that trigger acute inflammation in the air pouch inflammation model. When zymosan was used, the number of exudate cells first increased at 4h-8h, followed by a drop at 12h-24h. While, the changes in number of leukocytes in peripheral blood and proportion of neutrophils in bone marrow have the opposite trend. Meanwhile, neutrophils released neutrophil extracellular traps (NETs) to clean zymosan particles. In contrast, the cell migration response to carrageenan increased during 4h to 24h, no obvious NETs were observed, and the number of leukocytes in peripheral blood increased and the proportion of neutrophils in bone marrow decreased slightly. This study indicated that although both zymosan and carrageenan are sterile irritants, the characteristics of the inflammatory response induced by each other were different. In the acute phase of inflammation, zymosan-stimulated neutrophils were mobilized, recruited, and engulfed, and then died by NETs. Carrageenan stimulated the production of cytokines/chemokines by neutrophils or macrophages, but did not lead to an obvious death by releasing NETs.

An analytical method for solving the maximum deformation responses of suspension bridges with short extended spans under live load

As flexible structures, suspension bridges usually undergo significant deformation under live load when carrying rail traffic, jeopardizing the safety of passing trains. In engineering practice, suspension bridges with short extended spans have been applied to reduce the beam-end rotation angle under live load. However, the extended span layout is still determined empirically, and optimizing the span layout of such bridges to reduce maximum deformation under live load requires further investigation. This study proposes an analytical method for solving the maximum deformation responses of suspension bridges with extended spans under live load based on deflection theory and the extreme value analysis method. The proposed method can accurately solve the deformation responses of suspension bridges with extended spans caused by live load. It also allows one to analyze the most unfavorable live load distribution and the maximum deformation under live load, avoiding the redundant search process of the finite element method. Furthermore, the proposed method is applied to analyze the influence of the extended span layout on the maximum deformation responses under live load. The results show that adequately increasing the length of the first extended span and decreasing the length of the second extended span is beneficial to reduce the maximum beam-end rotation angle and the maximum curvature of the main beam.

GDCP-YOLO: Enhancing Steel Surface Defect Detection Using Lightweight Machine Learning Approach

Surface imperfections in steel materials potentially degrade quality and performance, thereby escalating the risk of accidents in engineering applications. Manual inspection, while traditional, is laborious and lacks consistency. However, recent advancements in machine learning and computer vision have paved the way for automated steel defect detection, yielding superior accuracy and efficiency. This paper introduces an innovative deep learning model, GDCP-YOLO, devised for multi-category steel defect detection. We enhance the reference YOLOv8n architecture by incorporating adaptive receptive fields via the DCNV2 module and channel attention in C2f. These integrations aim to concentrate on valuable features and minimize parameters. We incorporate the efficient Faster Block and employ Ghost convolutions to generate more feature maps with reduced computation. These modifications streamline feature extraction, curtail redundant information processing, and boost detection accuracy and speed. Comparative trials on the NEU-DET dataset underscore the state-of-the-art performance of GDCP-YOLO. Ablation studies and generalization experiments reveal consistent performance across a variety of defect types. The optimized lightweight architecture facilitates real-time automated inspection without sacrificing accuracy, offering invaluable insights to further deep learning techniques for surface defect identification across manufacturing sectors.

In situ edge extraction enabled by reconfigurable van der Waals infrared photodetectors

At present, the widely used artificial intelligence image perception technology is composed of discrete detection and processing components, which is mismatching the demand for increasingly sizeable redundant image information processing. The transmission of non-critical information between components limits the efficiency of image perception systems. Inspired by the human vision system, which can extract the key features and reduce the transmission burden at the low-level detection end, we propose a symmetric structured mixed-dimensional n/p/n [n-molybdenum disulfide (MoS2)/p-germanium (Ge)/n-MoS2] interdigital van der Waals heterojunction infrared photodetector that can be tuned by bias voltage in response polarity and magnitude. The MoS2/Ge/MoS2 infrared photodetector has a bias-symmetric optoelectronic response and covers the detection band from visible to short-wave infrared. Through the joint detection of a pair of MoS2/Ge/MoS2 infrared photodetectors, we demonstrate the image in situ edge extraction at the detection end, which provides key features for high-level processing. This work has discovered the potential of retina-inspired infrared photodetectors on a 2D/3D integration platform, providing distinct opportunities for a neuromorphic visual perception hardware.

Auditors’ Categorization of Process Deviations

ABSTRACT In addressing control deficiencies, auditors increasingly rely on data analytics. Despite the need to align information presentation with auditors’ cognitive structures, scant scholarly attention is given to how auditors internally categorize process deviations. This study investigates experienced auditors’ categorization of 62 deviations, revealing three primary categories: missing, reordered, and duplicated activities. These insights inform the development of active-learning algorithms, aligning with auditors’ knowledge structures to mitigate redundant processing risks. Blindly adopting process management research outcomes, however, poses a risk to auditing quality, impacting both effectiveness and efficiency in risk assessment and control testing. This research highlights the importance of validating and aligning deviation categories with auditors’ nuanced interpretations to enhance audit tools’ efficacy.

USING THE LEAN MANUFACTURING METHODOLOGY TO IMPROVE THE QUALITY OF THE ENTERPRISE'S BUSINESS PROCESSES

Formulation of the problem. In the conditions of global competition, lean production acquires special importance. For enterprises all over the world, the only way to survive in the market is to produce goods and provide services that most fully satisfy the demands of consumers, while using advanced management methods and technologies. Lean manufacturing is one of the modern tools of production organization and management technology, which is used to improve the quality of products and services, the most complete satisfaction of consumer needs, while ensuring a careful attitude to the company's resources and their savings.
 The purpose of the study is the further development of methodological, methodical and practical approaches to the concept of Lean manufacturing to improve the quality of the enterprise's business processes.
 Research methods. The research used scientific methods, such as theoretical generalization, comparison, analysis, synthesis.
 Result. In the process of research, a comparison of traditional business management and the concept of lean production was made and revealed. that the concept of lean production has such advantages as the transition from centralized, vertical management to horizontal, involvement of all employees in the process, teamwork, focus on rational use of resources. The concept of lean production is aimed at eliminating the following main types of losses: overproduction of goods; waiting for raw materials, semi-finished products, necessary service information; unnecessary movement of materials; redundant processing steps arising due to errors in the design of business processes; excessive stocks and business processes leading to their occurrence; unjustified rotation of personnel in the work process; production of waste; underutilization of employees' creative potential.
 Conclusions. Six main elements and 52 components (requirements) of Lean manufacturing are defined based on the SAE J4000 standard for more successful application of lean manufacturing ideas in practice production to analyze the levels of implementation of lean manufacturing methodology components and elements. The main elements of the concept of lean manufacturing, the development and improvement of which should be focused on, are: company management and trust in it; enterprise personnel; information environment; chain "supplier – organization – consumer"; product; business processes. An important role in the concept of lean manufacturing is played by the 5S ordering system and the system of general maintenance of TPM equipment .
 Keywords: lean manufacturing; losses; flow of value creation; 5S ordering system; system of general maintenance of TPM equipment; SAE standard J4000.

MPQ-YOLO: Ultra low mixed-precision quantization of YOLO for edge devices deployment

You Only Look Once (YOLO), known for its real-time performance and outstanding accuracy, has emerged as a prominent framework for object detection tasks. However, deploying YOLO on resource-constrained edge devices poses challenges due to its substantial memory requirements. In this paper, we propose MPQ-YOLO, an ultra-low mixed-precision quantization framework designed for edge device deployment. The core idea is to integrate 1-bit Backbone quantization and 4-bit Head quantization with dedicated training techniques. Specifically, we analyze the effect of numerical distribution on the performance of binary neural networks (BNNs), and based on this, we design a backbone with only 1-bit convolution. Then, we introduce a trainable scale and Progressive Network Quantization (PNQ) training strategy to bridge the Backbone and Head for end-to-end quantization training. The former is applied to both weights and activations within the 4-bit Head, enabling effective gradient propagation. The latter mitigates oscillation caused by mixed precision training, promoting smoother training and faster model convergence. Extensive experiments on VOC and COCO datasets demonstrate that MPQ-YOLO achieves a good trade-off between model compression and detection performance. Specifically, compared to the full-precision model, MPQ-YOLO achieves compression of up to 16.3× and 14.2× in terms of computational complexity and model size, respectively, while maintaining relatively high detection accuracy, i.e., 74.7% on VOC and 51.5% on COCO. To the best of our knowledge, MPQ-YOLO is the first YOLO framework with dual low mixed-precision quantization. Moreover, compared to the existing layer-wise mixed-precision quantization methods which cause redundant data processing and massive data movement, MPQ-YOLO offers a more hardware-designer-friendly and straightforward solution through efficient resource utilization and reuse.

Redundant Topology Design Method for Schedule Traffic Transmission Based on IEEE802.1CB Standard: Optimizing NW-PSP Scheduling Performance

This study proposes a method for designing redundant topologies in time-sensitive networking to transmit scheduled traffic (ST) while considering fault tolerance. In general, redundant topologies enhance network fault tolerance and are used in safety-critical applications that demand high reliability. These applications place a premium on real-time network performance, necessitating the consideration of the quality of service for time-critical traffic during the redundant topology design process. The primary objective of this study is to optimize the flowspan, which pertained to the scheduling performance of the no-wait packet scheduling problem during the redundant topology design phase. Specifically, we introduced a methodology for designing topologies tailored to transmit ST using the IEEE802.1CB mechanism. Comparisons with existing topology design approaches revealed the ability of the proposed method to enhance network fault tolerance and yield a topology that reduced flowspan of ST, thereby freeing up bandwidth for time-critical traffic.

Research on Malware Detection Technology for Mobile Terminals Based on API Call Sequence

With the development of the Internet, the types and quantities of malware have grown rapidly, and how to identify unknown malware is becoming a new challenge. The traditional malware detection method based on fixed features is becoming more and more difficult. In order to improve detection accuracy and efficiency for mobile terminals, this paper proposed a malware detection method for mobile terminals based on application programming interface (API) call sequence, which was characterized by the API call sequence and used a series of feature preprocessing techniques to remove redundant processing of the API call sequence. Finally, the recurrent neural network method (RNN) was used to build the model and perform detection and verification. Furthermore, this paper constructed a malware detection model based on a two-way recurrent neural network and used the two-way long short-term memory network model (LSTM) to train the data set containing 5986 malware samples and 5065 benign software samples to obtain the final detection model and its parameters. Finally, the feature vector of the APK file to be detected was passed into the model and obtained the detection results. The experimental results indicated that the detection accuracy of this method can reach 93.68%.

Decoding foveal word recognition: the role of interhemispheric inhibition in bilateral hemispheric processing

Extant research has largely favored the Split Fovea Theory (SFT) over the Bilateral Projection Theory (BPT) in the context of foveal word recognition. SFT posits that during foveal fixation, letters in the left and right visual fields are projected to their respective contralateral hemispheres, thereby facilitating a division of labor across the bilateral hemispheres. This division may serve as a regulatory mechanism to mitigate redundant processing in both hemispheres. The present investigation conducted two experiments utilizing Korean visual words to explore whether this hemispheric division in foveal word recognition is a strategy to circumvent potential interhemispheric inhibition arising from duplicated processing. Experiment 1 established the suitability of Korean visual words for studies involving both unilateral and bilateral presentations. Experiment 2 revealed that the split presentation of a word elicited greater accuracy compared to its identical presentation in the bilateral visual fields. These findings lend credence to the notion that interhemispheric inhibition may drive the hemispheres to engage in divided labor, thereby reducing processing redundancy in foveal word recognition.

Towards comprehensive understanding of pedestrians for autonomous driving: Efficient multi-task-learning-based pedestrian detection, tracking and attribute recognition

Safe interaction with pedestrians is important for autonomous vehicles in urban environments. Comprehensive understanding of pedestrian behavior is a challenge for autonomous driving, as different pedestrian attributes have different characteristics and there is not yet a dataset providing all attribute labels. In this paper, we design an efficient multi-task model for pedestrian detection, tracking and attribute recognition to fully mine information in multiple tasks and multiple datasets to reduce redundant image processing computation. Multiple training datasets are utilized to take full advantage of the available partially-labeled data. The model consists of two stages. The first stage is to detect and track pedestrians, using only low-resolution images to save computational resources. Accurate pedestrian detection is achieved in the bird's-eye-view (BEV), while pedestrian tracking is implemented on the basis of 3D bounding boxes. The second stage is image-based multi-attribute detection. This stage uses the detection and tracking results and some pre-processed information from the first stage along with high-resolution images. Multi-attributes of the pedestrian are identified, such as pedestrian's pose, age, gender, and motion direction. With a carefully-designed multi-task system that shares necessary feature maps among multiple tasks and reuses computation results, we can solve pedestrian detection, tracking, and multi-attribute recognition problems in an efficient framework. This leads to more accurate decisions and more efficient computations. The innovation of the model is to accomplish a simultaneous detection of multiple pedestrian attributes for autonomous driving, while fully sharing the features in the model to avoid repeated computation, as well as utilizing multiple partially-labeled datasets for the pedestrian detection, tracking and attribute recognition model training.

Control Scheme for a Quadratic-Based Step-Down On-Board DC/DC Converter to Be Used in Hybrid Electric Vehicles

This paper discusses a switching regulator with a quadratic-based step-down DC/DC converter designed using a reduced redundant power processing principle. This converter benefits from a non-complex topology, which aims to process energy more efficiently. The switching regulator can be used on-board in hybrid electric vehicles and has a non-pulsating input current, which is suitable to process energy from lithium-ion batteries. Design expressions and steady-state operating conditions are given for the converter. Models are obtained to design and implement a two-loop controller. The simplicity of this approach is of significant value in a method of design-oriented analysis, in which the analytic results can be used to make design choices. Loop-shaping techniques are used to design a robust controller to regulate the output voltage of the proposed converter under voltage variations on the lithium-ion batteries and the changes in load current requirements, as well as parameter uncertainties of the converter. The design conditions for the gains of the controller are derived. The models of the converter are verified through experimental results for a 300 W prototype and used to design a robust controller. Finally, closed-loop time and frequency domain tests are given to show the performance of the switching regulator.

Low-Cost Solution for Adding Safety Functions to Programmable Logic Controllers (PLCs)

Increasing requirements in automation and production make control systems more complex and vulnerable to failure. Breakdowns can cause delays in production, material damage, and above all, work-related accidents. For this reason, directives and legislation have been created at the country, European Union, and global levels that define the essential safety and requirements of industrial equipment. Guidelines must be observed by those involved in the design, supply, purchase, or use of industrial equipment in the European Union and several countries outside the European Union. Some guidelines (CAT, SIL) are created to ensure their safe operation in case of failure of both the hardware and the software. For a reliable operation of a fail safety system together with a system operating at SIL2 or SIL3, it must have Safety hardware and software. Industrial equipment manufacturers are incorporating security features into a variety of devices. Depending on the Safety Integrity Level (SIL) requirements, these features can be used during the design phase to increase safety in the event of failures or malfunctions. With proper design, the process as well as its environment (including people) can be protected by entering a controlled safe state. Manufacturers have approached this problem in a number of ways, including adding redundant Central Processing Units (CPUs), using special hardware to interface input and output signals, and developing secure network protocols for communication. Unfortunately, these features cannot be added to existing machines, at least without upgrading some hardware. As the associated costs lead to slower adoption, manufacturers rely on previous work to support certain security features, notably CPU debugging. This is implemented in the form of software libraries that operate at a low level (logic gate), designed to run on older hardware (PLC) so that they can offer an increased level of security. This study analyzes the required guidelines and legislation that must be followed for the safe operation of a production unit. It describes the mechanisms that basic and specialized PLC systems utilize to ensure the safety of an automation system. The authors have developed algorithms to record behavior measurements of electronic equipment. By further analyzing the results, it is concluded that basic equipment can be reused in these systems to provide safety functions, at the same time conserving both cost and time.

Biochemical characterization of an acid-thermostable glucoamylase from Aspergillus japonicus with potential application in the paper bio-deinking.

Aspergillus species have been highlighted in enzyme production looking for industrial applications, notably, amylases are one of the most interesting enzymes. They are capable of hydrolyzing α-glycosidic linkages of starch and widely used in industrial processes to produce ethanol, glucose, and fructose syrup as well as in the textiles, detergents, and paper industries applications. In this context, this work aimed at the biochemical characterization of the glucoamylase from Aspergillus japonicus and its application in the bio-bleaching process of recycled paper. The optimum temperature and pH for the glucoamylase assay were standardized as 50°C and 5.5. After 1 h of incubation, glucoamylase retained 90% of its activity at 30-50°C. It also kept 70% of its activity in the pH range of 4.0-6.5 after an hour of incubation. The enzyme led to an increase of 30% in the relative whiteness of 10 dry grams of sulfite paper and magazine paper when applied along with commercial cellulase and 10 mM MnCl2 . In addition, after the treatments, the glucoamylase recovered activity was 30%-32%, which indicates a prolonged availability of the enzyme and can considerably curtail the redundant downstream process of the recycled paper bio-bleaching. Thus, the glucoamylase from A. japonicus has a significant role in bio-bleaching recycled paper, reducing the necessity of hard chemicals, and improving the industrial process in an interesting economic and ecological mode.

IoVT-based efficient solution for optimal active smart camera selection in a tracking mission

Nowadays, to ensure people’s and property’s safety, cities have dozens of cameras installed on almost every corner. However, the haphazard deployment of these cameras often results in overlapping fields of view, leading to redundant processing of the same events by multiple cameras. This situation requires more power, huge bandwidth, and more storage capacity for collecting, processing and distributing multimedia data over the network. To overcome these challenges, we propose a novel Internet of Video Things-based system that efficiently reduces the number of cameras involved in processing redundant events. The proposed system operates through a two-step process. Firstly, a leader is elected from a group of cameras based on the degree of overlap in their fields of view, serving as the coordinator. Secondly, the system intelligently selects suitable assistants from neighboring cameras to complement the leader’s vision when it becomes inadequate for effective object tracking. Consequently, only the leader and its assistants remain active, while other neighboring cameras remain in an inactive state. The evaluation demonstrates that our IoVT-based solution offers an efficient approach by minimizing redundancy, optimizing resource utilization and overcoming the challenges posed by the random installation of cameras in urban environments.

A Fast Gradient Iterative Affine Motion Estimation Algorithm Based on Edge Detection for Versatile Video Coding

In the Versatile Video Coding (VVC) standard, affine motion models have been applied to enhance the resolution of complex motion patterns. However, due to the high computational complexity involved in affine motion estimation, real-time video processing applications face significant challenges. This paper focuses on optimizing affine motion estimation algorithms in the VVC environment and proposes a fast gradient iterative algorithm based on edge detection for efficient computation. Firstly, we establish judging conditions during the construction of affine motion candidate lists to streamline the redundant judging process. Secondly, we employ the Canny edge detection method for gradient assessment in the affine motion estimation process, thereby enhancing the iteration speed of affine motion vectors. The experimentalresults show that the encoding time of the affine motion estimation algorithm is about 15–35% lower than the overall encoding time of the anchor algorithm encoder, the average encoding time of the affine motion estimation part of the inter-frame prediction part is reduced by 24.79%, and the peak signal-to-noise ratio (PSNR) is only reduced by 0.04.