摘要 :
With the rapid development of the Internet technologies, the "Internet+" has become an important driving force for the transformation and upgrading of various industries. Driven by the new situations of the "Internet+", the big da...
展开
With the rapid development of the Internet technologies, the "Internet+" has become an important driving force for the transformation and upgrading of various industries. Driven by the new situations of the "Internet+", the big data has been rapidly infiltrated and applied to many fields and industries, playing a huge role in improving the working efficiency and creating the business values for enterprises. This article first describes the characteristics of the "Internet+" and the big data, and then explores the application of the big data in several main industries. On this basis, it explores the innovation of the Internet + Financial mode based on the big data, and the big data provides the environment and space for the financial model innovation. Finally, the paper discusses the innovation strategies for the financial model under the background of the big data.
收起
摘要 :
This paper gives the preprocessing technique research which includes odd-even calibration, bad pixels compensation and non-uniformity calibration for the high sensitivity push-broom long wave infrared camera. The noise equivalent ...
展开
This paper gives the preprocessing technique research which includes odd-even calibration, bad pixels compensation and non-uniformity calibration for the high sensitivity push-broom long wave infrared camera. The noise equivalent temperature difference (NETD) of the infrared camera has achieved less than 10 mK. The paper analyzes high sensitive imaging effects by using different preprocessing standards according to the test data, also analyzes the effects on imaging quality when using different preprocessing methods. At last the paper gives the suitable image preprocessing methods for the high sensitive infrared camera: the odd-even calibration method considering velocity to height ratio, the bad pixels compensation methods which are unique for temporal and spatial bad pixels and the non-uniformity calibration methods which combines the calibration based on black body testing data and the calibration based on real scene imaging data. It is validated that the residual nonuniformity of the infrared image can be reduced to 0.03% by using the preprocessing methods, which is satisfied for the requirement of high sensitive imaging.
收起
摘要 :
This paper gives the preprocessing technique research which includes odd-even calibration, bad pixels compensation and non-uniformity calibration for the high sensitivity push-broom long wave infrared camera. The noise equivalent ...
展开
This paper gives the preprocessing technique research which includes odd-even calibration, bad pixels compensation and non-uniformity calibration for the high sensitivity push-broom long wave infrared camera. The noise equivalent temperature difference (NETD) of the infrared camera has achieved less than 10 mK. The paper analyzes high sensitive imaging effects by using different preprocessing standards according to the test data, also analyzes the effects on imaging quality when using different preprocessing methods. At last the paper gives the suitable image preprocessing methods for the high sensitive infrared camera: the odd-even calibration method considering velocity to height ratio, the bad pixels compensation methods which are unique for temporal and spatial bad pixels and the non-uniformity calibration methods which combines the calibration based on black body testing data and the calibration based on real scene imaging data. It is validated that the residual nonuniformity of the infrared image can be reduced to 0.03% by using the preprocessing methods, which is satisfied for the requirement of high sensitive imaging.
收起
摘要 :
Multiple model-based engineering (MBE) frameworks have emerged to cover the many requirements for the engineering of avionics systems: from early requirement capture to the final system and embedded software generation, through re...
展开
Multiple model-based engineering (MBE) frameworks have emerged to cover the many requirements for the engineering of avionics systems: from early requirement capture to the final system and embedded software generation, through refinement and V&V activities. In this paper, we consider the SysML, AADL and FACE standards. They are promoted by different standardization bodies, with different objectives. We note they are often seen as competitive, while we argue it is the opposite: there is a potential for a synergistic coupling. To date, no complete open evaluation on the feasibility of such capability has been done. In this paper, we present one workflow that illustrates the joint use of SysML, AADL and FACE. We consider a basic flight control system to exercise the proposed process and gateways between the three notations. We use SCADE Architect by ANSYS that supports the three notations in a unified workbench to illustrate refinement scenarios from one notation to another, then synchronize with SCADE Suite to complete the software development.
收起
摘要 :
This paper gives the preprocessing technique research which includes odd-even calibration , bad pixels compensation and non-uniformity calibration for the high sensitivity push-broom long wave infrared camera. The noise equivalent...
展开
This paper gives the preprocessing technique research which includes odd-even calibration , bad pixels compensation and non-uniformity calibration for the high sensitivity push-broom long wave infrared camera. The noise equivalent temperature difference ( NETD ) of the infrared camera has achieved less than 10 mK. The paper analyzes high sensitive imaging effects by using different preprocessing standards according to the test data , also analyzes the effects on imaging quality when using different preprocessing methods. At last the paper gives the suitable image preprocessing methods for the high sensitive infrared camera:the odd-even calibration method considering velocity to height ratio,the bad pixels compensation methods which are unique for temporal and spatial bad pixels and the non-uniformity calibration methods which combines the calibration based on black body testing data and the calibration based on real scene imaging data. It is validated that the residual nonuniformity of the infrared image can be reduced to 0.03% by using the preprocessing methods, which is satisfied for the requirement of high sensitive imaging.
收起
摘要 :
Taking the advantage of the large mean free paths of the binder phases,we investigated the effect of Cr-V and Cr-V-Ln (Ln:rare earth) additives on the micro characteristics of the Co-based binder phases (the Cos) in WC-8.4wt％Co c...
展开
Taking the advantage of the large mean free paths of the binder phases,we investigated the effect of Cr-V and Cr-V-Ln (Ln:rare earth) additives on the micro characteristics of the Co-based binder phases (the Cos) in WC-8.4wt％Co cemented carbides with grain sizes larger than 5 μm and narrow two-phase carbon window.It included the crystal structure parameters,composition and morphology.To avoid the interference of WC on the analysis,a method of selective electrolysis corrosion of WC phase was employed.Based on the investigation of the residual Co skeletons,the following facts were established:1) the sole fcc structure; 2) the quite different solution behavior of V and Cr; 3) the significantly suppressed solid solubility of W and significantly increased solid solubility of W+Cr+V and 4) the formation of fine steps on the surfaces.A strong ability of Cr-V and Cr-V-Ln additives in the grain growth inhibition was observed even though an extra coarse WC raw material was used.The relationship among the solid solubility,lattice parameter and strain in/of the Cos,the magnetic saturation and the grain growth inhibition of the alloys were discussed.
收起
摘要 :
Taking the advantage of the large mean free paths of the binder phases,we investigated the effect of Cr-V and Cr-V-Ln (Ln:rare earth) additives on the micro characteristics of the Co-based binder phases (the Cos) in WC-8.4wt％Co c...
展开
Taking the advantage of the large mean free paths of the binder phases,we investigated the effect of Cr-V and Cr-V-Ln (Ln:rare earth) additives on the micro characteristics of the Co-based binder phases (the Cos) in WC-8.4wt％Co cemented carbides with grain sizes larger than 5 μm and narrow two-phase carbon window.It included the crystal structure parameters,composition and morphology.To avoid the interference of WC on the analysis,a method of selective electrolysis corrosion of WC phase was employed.Based on the investigation of the residual Co skeletons,the following facts were established:1) the sole fcc structure; 2) the quite different solution behavior of V and Cr; 3) the significantly suppressed solid solubility of W and significantly increased solid solubility of W+Cr+V and 4) the formation of fine steps on the surfaces.A strong ability of Cr-V and Cr-V-Ln additives in the grain growth inhibition was observed even though an extra coarse WC raw material was used.The relationship among the solid solubility,lattice parameter and strain in/of the Cos,the magnetic saturation and the grain growth inhibition of the alloys were discussed.
收起
摘要 :
In this paper, we present a coding framework for deep convolutional neural network compression. Our approach utilizes the classical coding theories and formulates the compression of deep convolutional neural networks as a rate-dis...
展开
In this paper, we present a coding framework for deep convolutional neural network compression. Our approach utilizes the classical coding theories and formulates the compression of deep convolutional neural networks as a rate-distortion optimization problem. We incorporate three coding ingredients in the coding framework, including bit allocation, dead zone quantization, and Tunstall coding, to improve the rate-distortion frontier without noticeable system-level overhead introduced. Experimental results show that our approach achieves state-of-the-art results on various deep convolutional neural networks and obtains considerable speedup on two deep learning accelerators. Specifically, our approach achieves 20× compression ratio on ResNet-18, ResNet-34, and ResNet-50, and 10× compression ratio on the compact already model MobileNet-v2, without hurting the accuracy. We then examine the system level impact of our approach when deploying the compressed models to hardware platforms. Hardware simulation results show that our approach obtains up to 4.3× and 2.8× inference speedup on state-of-the-art deep learning accelerators TPU and Eyeriss, respectively.
收起
原文获取