Limited Computational Resources Research Articles

Hierarchical aerial offload computing algorithm based on the Stackelberg-evolutionary game model

Aerial access networks have been envisioned as a promising 6 G technology to enhance the service experience in underserved areas where terrestrial base stations do not exist. In such scenarios, a hierarchical model of high-altitude platforms (HAPs) and unmanned aerial vehicles (UAVs) is considered to provide aerial computing services for ground Internet of Things (IoT) devices. In this study, we investigate a hierarchical aerial computing system to optimally orchestrate the limited computation resources in both HAPs and UAVs. For offloading services, we formulate a joint resource allocation problem to maximize service satisfaction for terrestrial IoT devices. To solve this problem, we employ the ideas of game theory with centralized decision and decentralized execution. Through the Stackelberg-evolutionary game model, the HAP works as a leader, and selects its price strategy based on the evolutionary learning process. As followers, individual UAVs make decisions to partially offload their computing tasks by considering different objectives. According to the interactive control paradigm, our proposed method can get reciprocal advantages for HAPs, UAVs, and ground IoT devices while adaptively handling dynamic aerial network conditions. Finally, extensive simulation results verify the efficiency of our proposed algorithm to increase the usability of edge servers’ computational resources. Compared with other existing state-of-the-art aerial network offloading protocols, we can improve the profits of system throughput, resource usability and UAV fairness up to 10 %, 10 %, and 15 %, respectively.

Object-Oriented and Visual-Based Localization in Urban Environments.

In visual-based localization, prior research falls short in addressing challenges for the Internet of Things with limited computational resources. The dominant state-of-the-art models are based on separate feature extractors and descriptors without consideration of the constraints of small hardware, the issue of inconsistent image scale, or the presence of multi-objects. We introduce "OOPose", a real-time object-oriented pose estimation framework that leverages dense features from off-the-shelf object detection neural networks. It balances between pixel-matching accuracy and processing speed, enhancing overall performance. When input images share a comparable set of features, their matching accuracy is substantially heightened, while the reduction in image size facilitates faster processing but may compromise accuracy. OOPose resizes both the original library and cropped query object images to a width of 416 pixels. This adjustment results in a 2.4-fold improvement in pose accuracy and an 8.6-fold increase in processing speed. Moreover, OOPose eliminates the need for traditional sparse point extraction and description processes by capitalizing on dense network backbone features and selecting the detected query objects and sources of object library images, ensuring not only 1.3 times more accurate results but also three times greater stability compared to real-time sparse ORB matching algorithms. Beyond enhancements, we demonstrated the feasibility of OOPose in an autonomous mobile robot, enabling self-localization with a single camera at 10 FPS on a single CPU. It proves the cost-effectiveness and real-world applicability of OOPose for small embedded devices, setting the stage for potential markets and providing end-users with distinct advantages.

Reservoir Computing Using Measurement-Controlled Quantum Dynamics

Physical reservoir computing (RC) is a machine learning algorithm that employs the dynamics of a physical system to forecast highly nonlinear and chaotic phenomena. In this paper, we introduce a quantum RC system that employs the dynamics of a probed atom in a cavity. The atom experiences coherent driving at a particular rate, leading to a measurement-controlled quantum evolution. The proposed quantum reservoir can make fast and reliable forecasts using a small number of artificial neurons compared with the traditional RC algorithm. We theoretically validate the operation of the reservoir, demonstrating its potential to be used in error-tolerant applications, where approximate computing approaches may be used to make feasible forecasts in conditions of limited computational and energy resources.

WS-SSD: Achieving faster 3D object detection for autonomous driving via weighted point cloud sampling

Due to the limited computational resources of the onboard computing devices of autonomous vehicles, the development of lightweight 3D object detectors is essential. Point-based detectors that progressively sample raw point clouds reduce numerous redundant computations and facilitate the implementation of high-speed 3D object detectors. The farthest point sampling based on Euclidean distance (D-FPS) is frequently utilized in point-based 3D detectors for point cloud sampling to reduce computation overhead. D-FPS can ensure uniform point sampling, covering the entire point cloud space as much as possible. The ratio of foreground and background points does not change significantly in the sampling results; hence, foreground objects do not receive sufficient attention. In addition, the number of road reflection points is large, and these points are close to foreground objects, reducing the sampling accuracy for faint objects. We propose weighted farthest point sampling based on Euclidean distance (W-DFPS), which selectively discards some road reflection points in the point sampling process, thus increasing the weight of foreground points in sampling results. It reduces the likelihood of faint objects being lost in the sampling results, such that a small number of sampling points can also cover most foreground objects. W-DFPS replaces D-FPS in the single-stage detector based on instance-aware (IA-SSD), and the network structure is slightly modified, namely the single-stage detector based on weighted sampling (WS-SSD). We evaluate WS-SSD in the KITTI dataset with a single A100 GPU. When the number of sampling points for the first module of WS-SSD is consistent with IA-SSD, the pedestrian detection accuracy is improved by an average of 4.06 compared to IA-SSD. Although the number of sampling points for the first module of WS-SSD is reduced to 25% of IA-SSD, the object detection accuracy remains competitive; it also achieves a state-of-the-art inference speed of 150.76 frames per second (FPS), which is a 46% improvement over IA-SSD.

SCSONet: spatial-channel synergistic optimization net for skin lesion segmentation

In the field of computer-assisted medical diagnosis, developing medical image segmentation models that are both accurate and capable of real-time operation under limited computational resources is crucial. Particularly for skin disease image segmentation, the construction of such lightweight models must balance computational cost and segmentation efficiency, especially in environments with limited computing power, memory, and storage. This study proposes a new lightweight network designed specifically for skin disease image segmentation, aimed at significantly reducing the number of parameters and floating-point operations while ensuring segmentation performance. The proposed ConvStem module, with full-dimensional attention, learns complementary attention weights across all four dimensions of the convolution kernel, effectively enhancing the recognition of irregularly shaped lesion areas, reducing the model’s parameter count and computational burden, thus promoting model lightweighting and performance improvement. The SCF Block reduces feature redundancy through spatial and channel feature fusion, significantly lowering parameter count while improving segmentation results. This paper validates the effectiveness and robustness of the proposed SCSONet on two public skin lesion segmentation datasets, demonstrating its low computational resource requirements. https://github.com/Haoyu1Chen/SCSONet.

The Effects of Speed and Delays on Test-Time Performance of End-to-End Self-Driving.

This study investigates the effects of speed variations and computational delays on the performance of end-to-end autonomous driving systems (ADS). Utilizing 1:10 scale mini-cars with limited computational resources, we demonstrate that different driving speeds significantly alter the task of the driving model, challenging the generalization capabilities of systems trained at a singular speed profile. Our findings reveal that models trained to drive at high speeds struggle with slower speeds and vice versa. Consequently, testing an ADS at an inappropriate speed can lead to misjudgments about its competence. Additionally, we explore the impact of computational delays, common in real-world deployments, on driving performance. We present a novel approach to counteract the effects of delays by adjusting the target labels in the training data, demonstrating improved resilience in models to handle computational delays effectively. This method, crucially, addresses the effects of delays rather than their causes and complements traditional delay minimization strategies. These insights are valuable for developing robust autonomous driving systems capable of adapting to varying speeds and delays in real-world scenarios.

Speed meets accuracy: Advanced deep learning for efficient Orientia tsutsugamushi bacteria assessment in RNAi screening

This study investigates the use of advanced computer vision techniques for assessing the severity of Orientia tsutsugamushi bacterial infectivity. It uses fluorescent scrub typhus images obtained from molecular screening, and addresses challenges posed by a complex and extensive image dataset, with limited computational resources. Our methodology integrates three key strategies within a deep learning framework: transitioning from instance segmentation (IS) models to an object detection model; reducing the model's backbone size; and employing lower-precision floating-point calculations. These approaches were systematically evaluated to strike an optimal balance between model accuracy and inference speed, crucial for effective bacterial infectivity assessment. A significant outcome is that the implementation of the Faster R-CNN architecture, with a shallow backbone and reduced precision, notably improves accuracy and reduces inference time in cell counting and infectivity assessment. This innovative approach successfully addresses the limitations of image processing techniques and IS models, effectively bridging the gap between sophisticated computational methods and modern molecular biology applications. The findings underscore the potential of this integrated approach to enhance the accuracy and efficiency of bacterial infectivity evaluations in molecular research.

A comparative study of large-scale and lightweight convolutional neural networks for ImageNet classification

In the field of convolution neural networks (CNNs), many impressive architectures have been published in recent years. These can be roughly divided into two groups: large-scale models and lightweight models. These large models are characterized by many trainable weights and complex network structures, offering them strong effectiveness in various computer vision tasks, and have become essential components of many modern visual recognition systems. These lightweight CNNs are designed to maintain high performance with limited memory and computational resources. They are highly efficient in terms of inference time and resource utilization, so that particularly suitable for mobile and edge computing devices. This work focuses on some prominent models based on the ImageNet database and explores the reasons for their frameworks success. By analyzing these models, a trend could be identified in the development of CNN models: Reasonably increasing the scale of the model and utilizing suitable frameworks can both improve accuracy and efficiency.

Cross-Parallel Attention and Efficient Match Transformer for Aerial Tracking

Visual object tracking is a key technology that is used in unmanned aerial vehicles (UAVs) to achieve autonomous navigation. In recent years, with the rapid development of deep learning, tracking algorithms based on Siamese neural networks have received widespread attention. However, because of complex and diverse tracking scenarios, as well as limited computational resources, most existing tracking algorithms struggle to ensure real-time stable operation while improving tracking performance. Therefore, studying efficient and fast-tracking frameworks, and enhancing the ability of algorithms to respond to complex scenarios has become crucial. Therefore, this paper proposes a cross-parallel attention and efficient match transformer for aerial tracking (SiamEMT). Firstly, we carefully designed the cross-parallel attention mechanism to encode global feature information and to achieve cross-dimensional interaction and feature correlation aggregation via parallel branches, highlighting feature saliency and reducing global redundancy information, as well as improving the tracking algorithm’s ability to distinguish between targets and backgrounds. Meanwhile, we implemented an efficient match transformer to achieve feature matching. This network utilizes parallel, lightweight, multi-head attention mechanisms to pass template information to the search region features, better matching the global similarity between the template and search regions, and improving the algorithm’s ability to perceive target location and feature information. Experiments on multiple drone public benchmark tests verified the accuracy and robustness of the proposed tracker in drone tracking scenarios. In addition, on the embedded artificial intelligence (AI) platform AGX Xavier, our algorithm achieved real-time tracking speed, indicating that our algorithm can be effectively applied to UAV tracking scenarios.

Cloud-edge collaborative computing offloading and resource allocation optimization in LEO satellite network

Satellite edge computing can provide communication and computing services to areas lacking ground network coverage, enhancing the computational capabilities of terrestrial users. However, due to the relatively limited computational resources available on satellites, cloud-edge collaborative computing in Low Earth Orbit (LEO) satellite communication scenarios emerges as a more optimal solution. In this paper, a cloud-edge collaborative offloading algorithm based on Deep Deterministic Policy Gradient (DDPG) algorithm is proposed to solve the offloading decision and resource allocation problems. The optimization problem is modeled as Markov decision process, and the comprehensive delay and energy consumption cost of computing tasks completed by ground users are optimized. Simulation results validate that the proposed algorithm has good convergence properties and lower comprehensive cost than processing all tasks locally and offloading all tasks to LEO satellites.

A hybrid neural network model based on transfer learning for Arabic sentiment analysis of customer satisfaction

AbstractSentiment analysis, a method used to classify textual content into positive, negative, or neutral sentiments, is commonly applied to data from social media platforms. Arabic, an official language of the United Nations, presents unique challenges for sentiment analysis due to its complex morphology and dialectal diversity. Compared to English, research on Arabic sentiment analysis is relatively scarce. Transfer learning, which applies the knowledge learned from one domain to another, can address the limitations of training time and computational resources. However, the development of transfer learning for Arabic sentiment analysis is still underdeveloped. In this study, we develop a new hybrid model, RNN‐BiLSTM, which merges recurrent neural networks (RNN) and bidirectional long short‐term memory (BiLSTM) networks. We used Arabic bidirectional encoder representations from transformers (AraBERT), a state‐of‐the‐art Arabic language pre‐trained transformer‐based model, to generate word‐embedding vectors. The RNN‐BiLSTM model integrates the strengths of RNN and BiLSTM, including the ability to learn sequential dependencies and bidirectional context. We trained the RNN‐BiLSTM model on the source domain, specifically the Arabic reviews dataset (ARD). The RNN‐BiLSTM model outperforms the RNN and BiLSTM models with default parameters, achieving an accuracy of 95.75%. We further applied transfer learning to the RNN‐BiLSTM model by fine‐tuning its parameters using random search. We compared the performance of the fine‐tuned RNN‐BiLSTM model with the RNN and BiLSTM models on two target domain datasets: ASTD and Aracust. The results showed that the fine‐tuned RNN‐BiLSTM model is more effective for transfer learning, achieving an accuracy of 95.44% and 96.19% on the ASTD and Aracust datasets, respectively.

Many-objective joint optimization of computation offloading and service caching in mobile edge computing

The computation offloading problem in mobile edge computing (MEC) has received a lot of attention, but service caching is also a research topic that cannot be ignored in MEC. Due to the limited resources available on the Edge Server (ES), a wise computation offloading and service caching policy must be formulated in order to maximize system offload efficiency. In this paper, a many-objective joint optimization computation offloading and service caching model (MaJOCOSC) is designed. The model takes into account the limited computing and storage resources of ES, the delay and energy consumption constraints of different types of tasks, and multiple processing modes of user tasks, and sets delay, energy consumption, task hit service rate, service cache balancing, and load balancing as the five optimization objectives of MaJOCOSC. Meanwhile, a non-dominated sorting genetic algorithm (NSGAIII-ASF&WD) based on achievement scalar function (ASF) and the k-nearest neighbor weighted distance mating selection strategy is proposed for better solving the model. The ASF ensures that the given strategy performs well for each objective value, and the k-nearest neighbor weighted distance provides the user with a diversity of strategies. Simulation results show that NSGAIII-ASF&WD can obtain better objective values when solving the model compared with other many-objective evolutionary algorithms, and a suitable computation offloading and service caching strategy is obtained.

Privacy-Preserving Serverless Edge Learning With Decentralized Small-Scale Mobile Data

In next-generation (i.e., 6G) networking systems, the data-driven approach will play an essential role, being an efficient tool for networking system management and bringing popular user applications. With those unprecedented and novel usages, existing frameworks fail to consider the complex nature of the next-generation networking system and consequently fail to be applied to future communication systems directly. Moreover, existing frameworks also fail to support popular privacy-preserving learning strategies efficiently by presenting special designs to respond to the resource-demanding nature of the aforementioned strategies. To fill this gap, this paper extends conventional serverless platforms with serverless edge learning architectures, providing a mature and efficient distributed training framework by fully exploiting limited wireless communication and edge computation resources in the considered networking system with the following three features. Firstly, this framework dynamically orchestrates resources among heterogeneous physical units to efficiently fulfill privacy-preserving learning objectives. The design jointly considers learning task requests and underlying infrastructure heterogeneity, including last-mile transmissions, computation abilities of edge and cloud computing centers, and loading status of infrastructure. Secondly, the proposed framework can easily work with data-driven approaches to improve network management efficiency, realizing AI for network promise of next-generation networking systems to provide efficient network automation. Lastly, to significantly reduce distributed training overheads, small-scale data training is proposed by integrating with a general, simple data classifier. This low-load enhancement can seamlessly work with various distributed deep models in the proposed framework to improve communications and computation efficiencies during the training phase. Based on the above innovations, open challenges, and future research directions encourage the research community to develop efficient privacy-preserving learning techniques.

Proficient Secret Key Generation using Derivates of the Transform Function for the Internet of Things

Secured data transfer over the internet of things enables communication devices is a very challenging task. Due to the limitation of computational resources, classical cryptography algorithms are vulnerable to cyber threats and loss of integrity and information. Recently, physical layer secret key generation has played a vital role in secured communication on the internet of things. The physical layer key generation approach provides an alternate means of information security using the signal reciprocity of wireless communication. This paper proposed proficient dynamic secret key generation using derivates of transform functions. The derivates of transform functions such as DCT, DWT, DWPT, and many more light transform functions the physical layer key generation employs RSSI channel parameters in conjunction with the discrete wavelet transform to generate low computational secret keys for the authentication of IoT-enabled communication devices. The proposed algorithm is very efficient and reduces the bit disagreement of the key. We use MATLAB software defined radios to conduct the experiment, and we then evaluate the effectiveness of our post-processing techniques. Our approach is helpful for safeguarding future IoT devices since it enhances the ability of targeted nodes to establish matching secret keys while reducing the risk of an eavesdropper. The experimental results of key generation methods suggest that the proposed algorithm is very efficient instead of DCT and DWT.

A novel anomaly detection model for the industrial Internet of Things using machine learning techniques

In recent decades, the pervasive integration of the Internet of Things (IoT) technologies has revolutionized various sectors, including industry 4.0, telecommunications, cloud computing, and healthcare systems. Industry 4.0 applications, characterized by real-time data exchange, increased reliance on automation, and limited computational resources at the edge, have reshaped global business dynamics, aiming to innovate business models through enhanced automation technologies. However, ensuring security in these environments remains a critical challenge, with real-time data streams introducing vulnerabilities to zero-day attacks and limited resources at the edge demanding efficient intrusion detection solutions. This study addresses this pressing need by proposing a novel intrusion detection model (IDS) specifically designed for Industry 4.0 environments. The proposed IDS leverages a Random Forest classifier with Principal Component Analysis (PCA) for feature selection. This approach addresses the challenges of real-time data processing and resource limitations while offering high accuracy. Based on the Bot-IoT dataset, the model achieves a competitive accuracy of 98.9% and a detection rate of 97.8%, outperforming conventional methods. This study demonstrates the effectiveness of the proposed IDS for securing Industry 4.0 ecosystems, offering valuable contributions to the field of cybersecurity.

A Multi-Task Network Based on Dual-Neck Structure for Autonomous Driving Perception.

A vision-based autonomous driving perception system necessitates the accomplishment of a suite of tasks, including vehicle detection, drivable area segmentation, and lane line segmentation. In light of the limited computational resources available, multi-task learning has emerged as the preeminent methodology for crafting such systems. In this article, we introduce a highly efficient end-to-end multi-task learning model that showcases promising performance on all fronts. Our approach entails the development of a reliable feature extraction network by introducing a feature extraction module called C2SPD. Moreover, to account for the disparities among various tasks, we propose a dual-neck architecture. Finally, we present an optimized design for the decoders of each task. Our model evinces strong performance on the demanding BDD100K dataset, attaining remarkable accuracy (Acc) in vehicle detection and superior precision in drivable area segmentation (mIoU). In addition, this is the first work that can process these three visual perception tasks simultaneously in real time on an embedded device Atlas 200I A2 and maintain excellent accuracy.

One-shot screening: Utilization of a two-dimensional convolutional neural network for automatic detection of left ventricular hypertrophy using electrocardiograms

Background and ObjectiveLeft ventricular hypertrophy (LVH) can impair ejection function and elevate the risk of heart failure. Therefore, early detection through screening is crucial. This study aimed to propose a novel method to enhance LVH detection using 12-lead electrocardiogram (ECG) waveforms with a two-dimensional (2D) convolutional neural network (CNN). MethodsUtilizing 42,127 pairs of ECG-transthoracic echocardiogram data, we pre-processed raw data into single-shot images derived from each ECG lead and conducted lead selection to optimize LVH diagnosis. Our proposed one-shot screening method, implemented during pre-processing, enables the superimposition of waveform source data of any length onto a single-frame image, thereby addressing the limitations of the one-dimensional (1D) approach. We developed a deep learning model with a 2D-CNN structure and machine learning models for LVH detection. To assess our method, we also compared our results with conventional ECG criteria and those of a prior study that used a 1D-CNN approach, utilizing the same dataset from the University of Tokyo Hospital for LVH diagnosis. ResultsFor LVH detection, the average area under the receiver operating characteristic curve (AUROC) was 0.916 for the 2D-CNN model, which was significantly higher than that obtained using logistic regression and random forest methods, as well as the two conventional ECG criteria (AUROC of 0.766, 0.790, 0.599, and 0.622, respectively). Incorporating additional metadata, such as ECG measurement data, further improved the average AUROC to 0.921. The model's performance remained stable across two different annotation criteria and demonstrated significant superiority over the performance of the 1D-CNN model used in a previous study (AUROC of 0.807). ConclusionsThis study introduces a robust and computationally efficient method that outperforms 1D-CNN models utilized in previous studies for LVH detection. Our method can transform waveforms of any length into fixed-size images and leverage the selected lead of the ECG, ensuring adaptability in environments with limited computational resources. The proposed method holds promise for integration into clinical practice as a tool for early diagnosis, potentially enhancing patient outcomes by facilitating earlier treatment and management.

Spatio-temporal neural network with handcrafted features for skeleton-based action recognition

The task of human action recognition (HAR) can be found in many computer vision practical applications. Various data modalities have been considered for solving this task, including joint-based skeletal representations which are suitable for real-time applications on platforms with limited computational resources. We propose a spatio-temporal neural network that uses handcrafted geometric features to classify human actions from video data. The proposed deep neural network architecture combines graph convolutional and temporal convolutional layers. The experiments performed on public HAR datasets show that our model obtains results similar to other state-of-the-art methods but has a lower inference time while offering the possibility to obtain an explanation for the classified action.

Deep Machine Learning of MobileNet, Efficient, and Inception Models

Today, specific convolution neural network (CNN) models assigned to specific tasks are often used. In this article, the authors explored three models: MobileNet, EfficientNetB0, and InceptionV3 combined. The authors were interested in investigating how quickly an artificial intelligence model can be taught with limited computer resources. Three types of training bases were investigated, starting with a simple base verifying five colours, then recognizing two different orthogonal elements, followed by more complex images from different families. This research aimed to demonstrate the capabilities of the models based on training base parameters such as the number of images and epoch types. Architectures proposed by the authors in these cases were chosen based on simulation studies conducted on a virtual machine with limited hardware parameters. The proposals present the advantages and disadvantages of the different models based on the TensorFlow and Keras libraries in the Jupiter environment based on the Python programming language. An artificial intelligence model with a combination of MobileNet, proposed by Siemens, and Efficient and Inception, selected by the authors, allows for further work to be conducted on image classification, but with limited computer resources for industrial implementation on a programmable logical controller (PLC). The study showed a 90% success rate, with a learning time of 180 s.

CrossPrune: Cooperative pruning for camera–LiDAR fused perception models of autonomous driving

Deep neural network pruning is effective in enabling high-performance perception models to be deployed on autonomous driving platforms with limited computation and memory resources. With their rapid development, state-of-the-art autonomous driving perception (ADP) models have advocated the use of multimodal sensors to extract diverse feature categories. However, existing pruning studies in the ADP area focus on single-modal models and neglect multimodal models. Compared with conventional pruning, multimodal pruning presents a new type of redundancy, namely modal-wise redundancy that is caused by multimodal branches extracting similar perception information. When a specific type of information is extracted by more than one modal branch, although this extraction is deemed essential from an individual single-modal standpoint, it becomes redundant when viewed from a modal-wise perspective. Therefore, modal branches must be handled cooperatively to eliminate modal-wise redundancy while concurrently preserving the original perception accuracy. Building on this, we propose CrossPrune, a modal cooperative pruning framework designed for camera–LiDAR fused perception in autonomous driving. The primary objective of CrossPrune is to effectively eliminate redundancy and achieve nondestructive pruning for multimodal ADP models. This was accomplished by approaching the problem as a multi-objective optimization task, encompassing both weight pruning and the restriction of feature distortions caused by pruning. Experiments conducted on the nuScenes and KITTI datasets demonstrated that CrossPrune attained superior pruning ratios while minimizing accuracy loss, surpassing the performance of the baselines. The key results indicated that the proposed CrossPrune achieved relative improvements of 9.6% in mAP and 11.5% in NDS under 89.8% pruning sparsity.