Graph Neural Networks(GNNs)have become a widely used tool for learning and analyzing data on graph structures,largely due to their ability to preserve graph structure and properties via graph representation ***,the ef...
详细信息
Graph Neural Networks(GNNs)have become a widely used tool for learning and analyzing data on graph structures,largely due to their ability to preserve graph structure and properties via graph representation ***,the effect of depth on the performance of GNNs,particularly isotropic and anisotropic models,remains an active area of *** study presents a comprehensive exploration of the impact of depth on GNNs,with a focus on the phenomena of over-smoothing and the bottleneck effect in deep graph neural *** research investigates the tradeoff between depth and performance,revealing that increasing depth can lead to over-smoothing and a decrease in performance due to the bottleneck *** also examine the impact of node degrees on classification accuracy,finding that nodes with low degrees can pose challenges for accurate *** experiments use several benchmark datasets and a range of evaluation metrics to compare isotropic and anisotropic GNNs of varying depths,also explore the scalability of these *** findings provide valuable insights into the design of deep GNNs and offer potential avenues for future research to improve their performance.
Pancreatic cancer's devastating impact and low survival rates call for improved detection methods. While Artificial Intelligence has shown remarkable progress, its increasing complexity has led to "black box&...
详细信息
In the analysis of drone aerial images, object detection tasks are particularly challenging, especially in the presence of complex terrain structures, extreme differences in target sizes, suboptimal shooting angles, a...
详细信息
In the analysis of drone aerial images, object detection tasks are particularly challenging, especially in the presence of complex terrain structures, extreme differences in target sizes, suboptimal shooting angles, and varying lighting conditions, all of which exacerbate the difficulty of recognition. In recent years, the DETR model based on the Transformer architecture has eliminated traditional post-processing steps such as NMS(Non-Maximum Suppression), thereby simplifying the object detection process and improving detection accuracy, which has garnered widespread attention in the academic community. However, DETR has limitations such as slow training convergence, difficulty in query optimization, and high computational costs, which hinder its application in practical fields. To address these issues, this paper proposes a new object detection model called OptiDETR. This model first employs a more efficient hybrid encoder to replace the traditional Transformer encoder. The new encoder significantly enhances feature processing capabilities through internal and cross-scale feature interaction and fusion logic. Secondly, an IoU (Intersection over Union) aware query selection mechanism is introduced. This mechanism adds IoU constraints during the training phase to provide higher-quality initial object queries for the decoder, significantly improving the decoding performance. Additionally, the OptiDETR model integrates SW-Block into the DETR decoder, leveraging the advantages of Swin Transformer in global context modeling and feature representation to further enhance the performance and efficiency of object detection. To tackle the problem of small object detection, this study innovatively employs the SAHI algorithm for data augmentation. Through a series of experiments, It achieved a significant performance improvement of more than two percentage points in the mAP (mean Average Precision) metric compared to current mainstream object detection models. Furthermore, ther
Alzheimer's disease is a common and complex brain disorder that primarily affects the elderly. Because it is progressing and has few effective therapies, it requires a thorough understanding of the condition;our s...
详细信息
Point cloud completion is crucial in point cloud processing, as it can repair and refine incomplete 3D data, ensuring more accurate models. However, current point cloud completion methods commonly face a challenge: th...
详细信息
The management of healthcare data has significantly benefited from the use of cloud-assisted MediVault for healthcare systems, which can offer patients efficient and convenient digital storage services for storin...
详细信息
Mobile banking security has witnessed significant R&D attention from both financial institutions and *** is due to the growing number of mobile baking applications and their reachability and usefulness to ***,thes...
详细信息
Mobile banking security has witnessed significant R&D attention from both financial institutions and *** is due to the growing number of mobile baking applications and their reachability and usefulness to ***,these applications are also attractive prey for cybercriminals,who use a variety of malware to steal personal banking *** literature in mobile banking security requiresmany permissions that are not necessary for the application’s intended security *** this context,this paper presents a novel efficient permission identification approach for securing mobile banking(MoBShield)to detect and prevent malware.A permission-based dataset is generated for mobile banking malware detection that consists large number of malicious adware apps and benign apps to use as training *** dataset is generated from 1650 malicious banking apps of the Canadian Institute of Cybersecurity,University of New Brunswick and benign apps from Google Play.A machine learning algorithm is used to determine whether amobile banking application ismalicious based on its permission ***,an eXplainable machine learning(XML)approach is developed to improve trust by explaining the reasoning behind the algorithm’s *** evaluation tests that the approach can effectively and practically identify mobile banking malware with high precision and reduced false ***,the adapted artificial neural networks(ANN),convolutional neural networks(CNN)and XML approaches achieve a higher accuracy of 99.7%and the adapted deep neural networks(DNN)approach achieves 99.6%accuracy in comparison with the state-of-the-art *** promising results position the proposed approach as a potential tool for real-world scenarios,offering a robustmeans of identifying and thwarting malware inmobile-based banking ***,MoBShield has the potential to significantly enhance the security and trustworthiness of
The Covid-19 pandemic has resulted in a permanent shift in individuals’ daily routines and driving behaviours, leading to an increase in remote work. There has also been an independent and parallel rise in the adopti...
详细信息
Ransomware is one of the most advanced malware which uses high computer resources and services to encrypt system data once it infects a system and causes large financial data losses to the organization and individuals...
详细信息
The scaler and scheduler of serverless system are the two cornerstones that ensure service quality and efficiency. However, existing scalers and schedulers are constrained by static thresholds, scaling latency, and si...
详细信息
暂无评论