With the advent of the era of big data,cloud computing,Internet of things,and other information industries continue to develop.There is an increasing amount of unstructured data such as pictures,audio,and video on the...With the advent of the era of big data,cloud computing,Internet of things,and other information industries continue to develop.There is an increasing amount of unstructured data such as pictures,audio,and video on the Internet.And the distributed object storage system has become the mainstream cloud storage solution.With the increasing number of distributed applications,data security in the distributed object storage system has become the focus.For the distributed object storage system,traditional defenses are means that fix discovered system vulnerabilities and backdoors by patching,or means to modify the corresponding structure and upgrade.However,these two kinds of means are hysteretic and hardly deal with unknown security threats.Based on mimic defense theory,this paper constructs the principle framework of the distributed object storage system and introduces the dynamic redundancy and heterogeneous function in the distributed object storage system architecture,which increases the attack cost,and greatly improves the security and availability of data.展开更多
Near-Earth Asteroids(NEA)impose potential major disaster to humanity.Planetary defense is an inevitable requirement for the survival of human civilization.In recent years,China has made rapid progress in planetary def...Near-Earth Asteroids(NEA)impose potential major disaster to humanity.Planetary defense is an inevitable requirement for the survival of human civilization.In recent years,China has made rapid progress in planetary defense research,which has won the attention of the government and attracted more and more scholars and organizations.This paper summarizes the research progress in planetary defense in China in recent years,including the fireball events in China,academic activities and policy planning,monitoring and warning technology,onorbit defense technology,impact hazard assessment,international cooperation and science popularization.展开更多
基于深度神经网络的目标检测技术已经广泛应用于各领域,然而,通过对抗补丁攻击在图像中添加局部扰动,以此来误导深度神经网络,对基于目标检测技术的视觉系统构成了严重威胁。针对这一问题,利用对抗补丁和图像背景的语义差异性,提出了一...基于深度神经网络的目标检测技术已经广泛应用于各领域,然而,通过对抗补丁攻击在图像中添加局部扰动,以此来误导深度神经网络,对基于目标检测技术的视觉系统构成了严重威胁。针对这一问题,利用对抗补丁和图像背景的语义差异性,提出了一种基于PatchTracker的对抗补丁防御算法,该算法由上游补丁检测器与下游数据增强模块组成。上游补丁检测器使用带有注意力机制的YOLOV5(you only look once-v5)确定对抗补丁所在位置,有助于提高对小尺度对抗补丁的检测精度;将检测区域用合适的像素值覆盖以抹除对抗补丁,上游补丁检测器不仅能够有效降低对抗样本的攻击性,而且不依赖大规模的训练数据;下游数据增强模块通过改进模型训练范式,提高下游目标检测器的鲁棒性;将抹除补丁后的图像输入经过数据增强的下游YOLOV5目标检测模型。在公开的TT100K交通标志数据集上进行了交叉验证,实验表明,与未采取防御措施相比,所提算法能够有效防御多种类型的通用对抗补丁攻击,在检测对抗补丁图像时的mAP(mean average precision)提高65%左右,有效地改善了小尺度对抗补丁的漏检情况。与现有算法比较,所提算法有效提高了神经网络在检测对抗样本时的准确率。此外,所提算法不涉及下游模型结构的修改,具有良好的兼容性。展开更多
基金National Keystone R&D Program of China(No.2017YFB0803204)Shenzhen Research Programs(JCYJ20170306092030521)+3 种基金the PCL Future Regional Network Facilities for Largescale Experiments and Applications(LZC0019)ZTE University Funding,Natural Science Foundation of China(NSFC)(No.61671001)GuangDong Prov.,R&D Key Program(No.2019B010137001)the Shenzhen Municipal Development and Reform Commission(Disciplinary Development Program for Data Science and Intelligent Computing).
文摘With the advent of the era of big data,cloud computing,Internet of things,and other information industries continue to develop.There is an increasing amount of unstructured data such as pictures,audio,and video on the Internet.And the distributed object storage system has become the mainstream cloud storage solution.With the increasing number of distributed applications,data security in the distributed object storage system has become the focus.For the distributed object storage system,traditional defenses are means that fix discovered system vulnerabilities and backdoors by patching,or means to modify the corresponding structure and upgrade.However,these two kinds of means are hysteretic and hardly deal with unknown security threats.Based on mimic defense theory,this paper constructs the principle framework of the distributed object storage system and introduces the dynamic redundancy and heterogeneous function in the distributed object storage system architecture,which increases the attack cost,and greatly improves the security and availability of data.
基金Supported by the Beijing Municipal Science and Technology Commission(Z181100002918004)the Strategic Priority Program on Space Science(XDA15014900)the Civil Aerospace Preliminary Research Project(KJSP2020020101,CAS)。
文摘Near-Earth Asteroids(NEA)impose potential major disaster to humanity.Planetary defense is an inevitable requirement for the survival of human civilization.In recent years,China has made rapid progress in planetary defense research,which has won the attention of the government and attracted more and more scholars and organizations.This paper summarizes the research progress in planetary defense in China in recent years,including the fireball events in China,academic activities and policy planning,monitoring and warning technology,onorbit defense technology,impact hazard assessment,international cooperation and science popularization.
文摘基于深度神经网络的目标检测技术已经广泛应用于各领域,然而,通过对抗补丁攻击在图像中添加局部扰动,以此来误导深度神经网络,对基于目标检测技术的视觉系统构成了严重威胁。针对这一问题,利用对抗补丁和图像背景的语义差异性,提出了一种基于PatchTracker的对抗补丁防御算法,该算法由上游补丁检测器与下游数据增强模块组成。上游补丁检测器使用带有注意力机制的YOLOV5(you only look once-v5)确定对抗补丁所在位置,有助于提高对小尺度对抗补丁的检测精度;将检测区域用合适的像素值覆盖以抹除对抗补丁,上游补丁检测器不仅能够有效降低对抗样本的攻击性,而且不依赖大规模的训练数据;下游数据增强模块通过改进模型训练范式,提高下游目标检测器的鲁棒性;将抹除补丁后的图像输入经过数据增强的下游YOLOV5目标检测模型。在公开的TT100K交通标志数据集上进行了交叉验证,实验表明,与未采取防御措施相比,所提算法能够有效防御多种类型的通用对抗补丁攻击,在检测对抗补丁图像时的mAP(mean average precision)提高65%左右,有效地改善了小尺度对抗补丁的漏检情况。与现有算法比较,所提算法有效提高了神经网络在检测对抗样本时的准确率。此外,所提算法不涉及下游模型结构的修改,具有良好的兼容性。
