Due to their detection capabilities and low cost,unmanned aerial vehicles(UAVs)are commonly used in search and rescue(SAR)operations.In a SAR mission,the UAV's height may change,causing objects to shrink or increa...Due to their detection capabilities and low cost,unmanned aerial vehicles(UAVs)are commonly used in search and rescue(SAR)operations.In a SAR mission,the UAV's height may change,causing objects to shrink or increase thus affecting generalization.This paper proposes a hybrid object detection(OD)technique that combines altitude-dependent local deep learning(DL)models,each one designed for a given flight altitude range.Seven cutting-edge OD models,including YOLOv4 and v5,Efficient Det,Detectron2,Mobile Net,and Faster R-CNN,were trained locally with YOLOv5 and Scaled YOLOv4 being the best performers in low and high-altitude images,respectively.The suggested hybrid strategy,which uses the best OD performers,outperformed well-known DL algorithms with 86.2%m AP on a public dataset.Computing efficiency and accuracy with images of varying resolutions were also explored.Dividing the fundamental detection issue into local subproblems that are treated separately by powerful OD networks might increase SAR capabilities.展开更多
Intrusion detection systems(IDS)can play a significant role in detecting security threats or malicious attacks that aim to steal information and/or corrupt network protocols.To deal with the dynamic and complex nature...Intrusion detection systems(IDS)can play a significant role in detecting security threats or malicious attacks that aim to steal information and/or corrupt network protocols.To deal with the dynamic and complex nature of cyber-attacks,advanced intelligent tools have been applied resulting into powerful and automated IDS that rely on the latest advances of machine learning(ML)and deep learning(DL).Most of the reported effort has been devoted on building complex ML/DL architectures adopting a brute force approach towards the maximization of their detection capacity.However,just a limited number of studies have focused on the identification or extraction of user-friendly risk indicators that could be easily used by security experts.Many papers have explored various dimensionality reduction algorithms,however a large number of selected features is still required to detect the attacks successfully,which humans cannot intuitively or immediately understand.To enhance user’s trust and understanding on data without sacrificing on accuracy,this paper contributes to the transformation of the available data collected by IDS into a single actionable and easy-to-understand risk indicator.To achieve this,a novel feature extraction pipeline was implemented consisting of the following components:(i)a fuzzy allocation scheme that transforms raw data to fuzzy class memberships,(ii)a novel modality transformation mechanism for converting feature vectors to images(Vec2im)and(iii)a dimensionality reduction module that makes use of Siamese convolutional neural networks that finally reduces the input data dimensionality into a 1-d feature space.The performance of the proposed methodology was validated with respect to detection accuracy,dimensionality reduction performance and execution time on the NSL-KDD dataset via a thorough comparative analysis that demonstrated its effectiveness(86.64%testing accuracy using only one feature)over a number of well-known feature selection(FS)and extraction techniques.The output of the proposed feature extraction pipeline could be potentially used by security experts as an indicator of malicious activity,whereas the generated images could be further utilized and/or integrated as a visual analytics tool in existing IDS.展开更多
基金supported by Horizon 2020 Framework Programme:[Grant Number 882897]。
文摘Due to their detection capabilities and low cost,unmanned aerial vehicles(UAVs)are commonly used in search and rescue(SAR)operations.In a SAR mission,the UAV's height may change,causing objects to shrink or increase thus affecting generalization.This paper proposes a hybrid object detection(OD)technique that combines altitude-dependent local deep learning(DL)models,each one designed for a given flight altitude range.Seven cutting-edge OD models,including YOLOv4 and v5,Efficient Det,Detectron2,Mobile Net,and Faster R-CNN,were trained locally with YOLOv5 and Scaled YOLOv4 being the best performers in low and high-altitude images,respectively.The suggested hybrid strategy,which uses the best OD performers,outperformed well-known DL algorithms with 86.2%m AP on a public dataset.Computing efficiency and accuracy with images of varying resolutions were also explored.Dividing the fundamental detection issue into local subproblems that are treated separately by powerful OD networks might increase SAR capabilities.
基金This work has received funding from the European Community’s H2020 Programme,under grant agreement Nr.826183(SPHINX).
文摘Intrusion detection systems(IDS)can play a significant role in detecting security threats or malicious attacks that aim to steal information and/or corrupt network protocols.To deal with the dynamic and complex nature of cyber-attacks,advanced intelligent tools have been applied resulting into powerful and automated IDS that rely on the latest advances of machine learning(ML)and deep learning(DL).Most of the reported effort has been devoted on building complex ML/DL architectures adopting a brute force approach towards the maximization of their detection capacity.However,just a limited number of studies have focused on the identification or extraction of user-friendly risk indicators that could be easily used by security experts.Many papers have explored various dimensionality reduction algorithms,however a large number of selected features is still required to detect the attacks successfully,which humans cannot intuitively or immediately understand.To enhance user’s trust and understanding on data without sacrificing on accuracy,this paper contributes to the transformation of the available data collected by IDS into a single actionable and easy-to-understand risk indicator.To achieve this,a novel feature extraction pipeline was implemented consisting of the following components:(i)a fuzzy allocation scheme that transforms raw data to fuzzy class memberships,(ii)a novel modality transformation mechanism for converting feature vectors to images(Vec2im)and(iii)a dimensionality reduction module that makes use of Siamese convolutional neural networks that finally reduces the input data dimensionality into a 1-d feature space.The performance of the proposed methodology was validated with respect to detection accuracy,dimensionality reduction performance and execution time on the NSL-KDD dataset via a thorough comparative analysis that demonstrated its effectiveness(86.64%testing accuracy using only one feature)over a number of well-known feature selection(FS)and extraction techniques.The output of the proposed feature extraction pipeline could be potentially used by security experts as an indicator of malicious activity,whereas the generated images could be further utilized and/or integrated as a visual analytics tool in existing IDS.
