In this paper an event-based operational interleaving semantics is proposed for real-time processes, for which action refinement and a denotational true concurrency semantics are developed and defined in terms of time...In this paper an event-based operational interleaving semantics is proposed for real-time processes, for which action refinement and a denotational true concurrency semantics are developed and defined in terms of timed event structures. The authors characterize the timed event traces that are generated by the operational semantics in a denotational way, and show that this operational semantics is consistent with the denotational semantics in the sense that they generate the same set of timed event traces, thereby eliminating the gap between the true concurrency and interleaving semantics. Keywords action refinement - real-time process algebra - semantics - timed event structure - formal method This work was supported by the National Natural Science Foundation of China (Grant No. 60373113) and the “Hundred-Talent Program” of Chinese Academy of Sciences.Xiu-Li Sun was born in 1975. She received her B.S. degree in 1998 and M.S. degree in 2002 from Taiyuan University of Technology, Shanxi. She is studying in the Institute of Computer Application, Chengdu for her doctorate.Wen-Ying Zhang was born in 1972. Now he is a Ph.D. candidate of Computer Application, the Chinese Academy of Sciences. His current research interests include formal verification, digital watermarking and pattern recognition.Jin-Zhao Wu was born in 1965. He obtained his Ph.D. degree in 1994 from the Institute of System Science, CAS. From 1994 to 1999 he was a postdoctoral researcher. His research interests include formal specification and verification, automatic reasoning, logic programming.展开更多
针对小样本语义分割中同类别支持图像与查询图像存在外观差异较大的问题,提出融合高斯过程的自支持匹配小样本语义分割模型。提出的模型在自支持匹配小样本语义分割模型的基础上,首先融入高斯过程,对分布在深层特征空间上的复杂外观进...针对小样本语义分割中同类别支持图像与查询图像存在外观差异较大的问题,提出融合高斯过程的自支持匹配小样本语义分割模型。提出的模型在自支持匹配小样本语义分割模型的基础上,首先融入高斯过程,对分布在深层特征空间上的复杂外观进行建模,捕获更多空间细节信息来表示数据分布;随后设计特征增强模块,在空间层对支持特征与查询特征进行信息交互,在通道层进行注意力加权,进一步增强相同类之间的全局相似性,捕获更多目标类别信息;最后利用Gram矩阵量化支持图像和查询图像之间外观差异的大小,从而融合原型匹配的结果,产生更准确的分割图像。实验结果表明:与现有方法相比,所提模型在更强的主干网络下具有较好的分割结果和更少的参数量,在5-shot的设定下,所提模型在PASCAL−5i数据集上平均交并比(mean Intersection over Union,mIoU)达到最优值,提升了0.4%;在COCO−20i数据集上的子集mIoU取得最优值,分别提升了2.2%和1.0%,表明该模型的有效性和先进性。展开更多
岩土地层结构是影响工程稳定与安全的重要因素,是地质及岩土工程勘测的主要内容。研发了一种地层地质界面识别系统,阐述了其基本组成和工作原理,建立了钻进过程识别码及岩土地层结构实时识别方法;通过花岗岩地基钻进试验研究,确定了岩...岩土地层结构是影响工程稳定与安全的重要因素,是地质及岩土工程勘测的主要内容。研发了一种地层地质界面识别系统,阐述了其基本组成和工作原理,建立了钻进过程识别码及岩土地层结构实时识别方法;通过花岗岩地基钻进试验研究,确定了岩体风化等级以及土层/散体、碎裂岩体及整体岩体结构的比能值范围,建立了金刚石钻进比能的地层结构分级标准,比能值分别是土层及散体为0.0017~0.6955 k N·m/cm3,碎裂岩体为0.0258~0.9798 k N·m/cm3,整体岩体为0.0018~2.4935 k N·m/cm3。研究结果表明,该方法与岩土勘测结果一致,可用于地层地质及岩土地层结构的实时识别,为岩土工程智能勘测提供了新的途径。展开更多
文摘In this paper an event-based operational interleaving semantics is proposed for real-time processes, for which action refinement and a denotational true concurrency semantics are developed and defined in terms of timed event structures. The authors characterize the timed event traces that are generated by the operational semantics in a denotational way, and show that this operational semantics is consistent with the denotational semantics in the sense that they generate the same set of timed event traces, thereby eliminating the gap between the true concurrency and interleaving semantics. Keywords action refinement - real-time process algebra - semantics - timed event structure - formal method This work was supported by the National Natural Science Foundation of China (Grant No. 60373113) and the “Hundred-Talent Program” of Chinese Academy of Sciences.Xiu-Li Sun was born in 1975. She received her B.S. degree in 1998 and M.S. degree in 2002 from Taiyuan University of Technology, Shanxi. She is studying in the Institute of Computer Application, Chengdu for her doctorate.Wen-Ying Zhang was born in 1972. Now he is a Ph.D. candidate of Computer Application, the Chinese Academy of Sciences. His current research interests include formal verification, digital watermarking and pattern recognition.Jin-Zhao Wu was born in 1965. He obtained his Ph.D. degree in 1994 from the Institute of System Science, CAS. From 1994 to 1999 he was a postdoctoral researcher. His research interests include formal specification and verification, automatic reasoning, logic programming.
文摘针对小样本语义分割中同类别支持图像与查询图像存在外观差异较大的问题,提出融合高斯过程的自支持匹配小样本语义分割模型。提出的模型在自支持匹配小样本语义分割模型的基础上,首先融入高斯过程,对分布在深层特征空间上的复杂外观进行建模,捕获更多空间细节信息来表示数据分布;随后设计特征增强模块,在空间层对支持特征与查询特征进行信息交互,在通道层进行注意力加权,进一步增强相同类之间的全局相似性,捕获更多目标类别信息;最后利用Gram矩阵量化支持图像和查询图像之间外观差异的大小,从而融合原型匹配的结果,产生更准确的分割图像。实验结果表明:与现有方法相比,所提模型在更强的主干网络下具有较好的分割结果和更少的参数量,在5-shot的设定下,所提模型在PASCAL−5i数据集上平均交并比(mean Intersection over Union,mIoU)达到最优值,提升了0.4%;在COCO−20i数据集上的子集mIoU取得最优值,分别提升了2.2%和1.0%,表明该模型的有效性和先进性。
文摘岩土地层结构是影响工程稳定与安全的重要因素,是地质及岩土工程勘测的主要内容。研发了一种地层地质界面识别系统,阐述了其基本组成和工作原理,建立了钻进过程识别码及岩土地层结构实时识别方法;通过花岗岩地基钻进试验研究,确定了岩体风化等级以及土层/散体、碎裂岩体及整体岩体结构的比能值范围,建立了金刚石钻进比能的地层结构分级标准,比能值分别是土层及散体为0.0017~0.6955 k N·m/cm3,碎裂岩体为0.0258~0.9798 k N·m/cm3,整体岩体为0.0018~2.4935 k N·m/cm3。研究结果表明,该方法与岩土勘测结果一致,可用于地层地质及岩土地层结构的实时识别,为岩土工程智能勘测提供了新的途径。
