The phase transformation characteristics of a high-strength TRIP-aided multiphase cold-rolled steel during continuous heating at different cooling rates were studied by means of dilatometry,and the critical temperatur...The phase transformation characteristics of a high-strength TRIP-aided multiphase cold-rolled steel during continuous heating at different cooling rates were studied by means of dilatometry,and the critical temperatures were also determined.The samples were fully austenitized at 1 050 ℃ and then cooled at different cooling rates ranging from 0.5 ℃/s to 100 ℃/s.The continuous cooling transformation （CCT） curves were obtained for the experimental steel.The experimental results showed that a high cooling rate depressed the formation of ferrite and pearlite and promoted the formation of balnite and martensite,leading to a higher hardness.A large amount of martensite in high-strength TRIP-aided multiphase cold-rolled steel can be obtained at cooling rates in excess of 50 ℃/s.The experimental results provide guidelines for cooling control and heat treatment in real steel production.展开更多
This paper analyzes the long-run effects and short-run effects of foreign aid on the domestic economy by using the Hamilton system and Laplace transform. It is found that an increase in the foreign aid has no long-run...This paper analyzes the long-run effects and short-run effects of foreign aid on the domestic economy by using the Hamilton system and Laplace transform. It is found that an increase in the foreign aid has no long-run effect on the foreigll borrowing, domestic capital accumulation and the foreign direct investment in the home country, but increases the steady-state consumption level the same amount. However, the short-run analysis presents that increasing foreign aid does not affect the initial consumptioll level and the initial consumption increase rate; but it affects the initial savings positively.展开更多
In this thesis, a strategy realizing the computer-aided detection (CAD) of the epileptic waves in EEG isintroduced. The expert criterion, continuous wavelet transformation, neural networks, and characteristic paramete...In this thesis, a strategy realizing the computer-aided detection (CAD) of the epileptic waves in EEG isintroduced. The expert criterion, continuous wavelet transformation, neural networks, and characteristic parametermeasuremente these modern signa1 processing weapons were synthesized togetLher to form a so-called multi-method.It was estimated that the advantages of all the powerful techniques could be exploited systematically. Therefore, theCAD’s capacities in the long-term monitoring, trCaAnent and control of epilepsy might be enhanced. In this strategy,the raw EEG signals were uniformed and the expelt criterion were applied to discard most of aItifacts in them at first,and then the signals were pre-processed by continuous wavelet transformation. Some characteristic parameters wereextracted from the raw signals and the pre-processed ones. Consequently groups of eighteen parameters were sent totrain or test BP networks. By applying this theme a correct-detection rate of 84.3% for spike and sharp waves, and88.9% for sPike and sharp slow waves were obtained. In the next step, some non-linear tools wtll also be equippedwith the CAD system.展开更多
A 3-Dimensional computer aided garment design (CAGD) system has been developed andimplemented on a high-performance workstation. We studied various approaches to the func-tional modelling of garment designs for the sy...A 3-Dimensional computer aided garment design (CAGD) system has been developed andimplemented on a high-performance workstation. We studied various approaches to the func-tional modelling of garment designs for the system. According to the characteristic data of a hu-man body, the models of human body and the garment are displayed on the screen, then we canmodify the garment with various styles and different sizes. The system can transform the 3-Dgarment to the 2-D pieces. The system has improved design efficiency. Various potential alterna-tives and improvement of the system have also been studied and explored.展开更多
This paper presents a general method for 2D/3D transformation, which can be efficiently used in three dimensional computer aided garment design. The method utilizes a uniform triangular spring_mass based deformable mo...This paper presents a general method for 2D/3D transformation, which can be efficiently used in three dimensional computer aided garment design. The method utilizes a uniform triangular spring_mass based deformable model. 2D to 3D transformation and 3D to 2D transformation both can be implemented on the same model. A general and efficient collision detection method is also briefly discussed in this paper.展开更多
目的经颅超声成像技术作为高效率、低成本且无创的诊断手段,已逐步应用于帕金森病患者认知功能障碍诊断。由于经颅超声图像信噪比低、成像质量差、目标组织复杂且相似度高,需要依赖专业医生手动检测。但是人工检测不仅费时费力,还可能...目的经颅超声成像技术作为高效率、低成本且无创的诊断手段,已逐步应用于帕金森病患者认知功能障碍诊断。由于经颅超声图像信噪比低、成像质量差、目标组织复杂且相似度高,需要依赖专业医生手动检测。但是人工检测不仅费时费力,还可能因为操作者的主观因素影响,造成检测结果出现差异性。针对这一问题,提出了一种基于Swin Transformer和多尺度深度特征融合的YOLO-SF-TV(YOLO network based on Swin Transformer and multiscale deep feature fusion for third ventricle)模型用于经颅超声图像三脑室检测,以提高临床检测准确率,辅助医生进行早期诊断。方法YOLO-SF-TV模型在YOLOv8的基础上使用基于窗口注意力的Swin Transformer作为模型特征提取网络,并引入空间金字塔池化合模块SPP-FCM(spatial pyramid pooling fast incorporating CSPNet and multiple attention mechanisms)扩大网络感受野,并增强多尺度特征融合能力。在网络的多尺度特征融合部分结合深度可分离卷积和多头注意力机制,提出了PAFPN-DM(path aggregation and feature pyramid network with depthwise separable convolution)模块,并对主干特征输出层增加多头注意力机制,以提高网络对不同尺度特征图中全局和局部重要信息的理解能力。同时,将传统卷积替换为深度可分离卷积模块,通过对每个通道单独卷积提高网络对不同通道的敏感性,以保证模型准确度的同时降低训练参数和难度,增强模型的泛化能力。结果在本文收集的经颅超声三脑室图像数据及对应标签的数据集上进行实验,并与典型的目标检测模型对比。实验结果表明,本文提出的YOLO-SF-TV在经颅超声三脑室目标上的平均精确度均值(mean average precision,mAP)达到98.69%,相比于YOLOv8提升了2.12%,与其他典型模型相比检测精度达到最优。结论本文提出的YOLO-SF-TV模型在经颅超声图像三脑室检测问题上表现优秀,SPP-FCM模块和PAFPN-DM模块可以增强模型检测能力,提高模型泛化性和鲁棒性。同时,本文制作的数据集将有助于推动经颅超声三脑室图像检测问题的研究。展开更多
文摘The phase transformation characteristics of a high-strength TRIP-aided multiphase cold-rolled steel during continuous heating at different cooling rates were studied by means of dilatometry,and the critical temperatures were also determined.The samples were fully austenitized at 1 050 ℃ and then cooled at different cooling rates ranging from 0.5 ℃/s to 100 ℃/s.The continuous cooling transformation （CCT） curves were obtained for the experimental steel.The experimental results showed that a high cooling rate depressed the formation of ferrite and pearlite and promoted the formation of balnite and martensite,leading to a higher hardness.A large amount of martensite in high-strength TRIP-aided multiphase cold-rolled steel can be obtained at cooling rates in excess of 50 ℃/s.The experimental results provide guidelines for cooling control and heat treatment in real steel production.
文摘This paper analyzes the long-run effects and short-run effects of foreign aid on the domestic economy by using the Hamilton system and Laplace transform. It is found that an increase in the foreign aid has no long-run effect on the foreigll borrowing, domestic capital accumulation and the foreign direct investment in the home country, but increases the steady-state consumption level the same amount. However, the short-run analysis presents that increasing foreign aid does not affect the initial consumptioll level and the initial consumption increase rate; but it affects the initial savings positively.
文摘In this thesis, a strategy realizing the computer-aided detection (CAD) of the epileptic waves in EEG isintroduced. The expert criterion, continuous wavelet transformation, neural networks, and characteristic parametermeasuremente these modern signa1 processing weapons were synthesized togetLher to form a so-called multi-method.It was estimated that the advantages of all the powerful techniques could be exploited systematically. Therefore, theCAD’s capacities in the long-term monitoring, trCaAnent and control of epilepsy might be enhanced. In this strategy,the raw EEG signals were uniformed and the expelt criterion were applied to discard most of aItifacts in them at first,and then the signals were pre-processed by continuous wavelet transformation. Some characteristic parameters wereextracted from the raw signals and the pre-processed ones. Consequently groups of eighteen parameters were sent totrain or test BP networks. By applying this theme a correct-detection rate of 84.3% for spike and sharp waves, and88.9% for sPike and sharp slow waves were obtained. In the next step, some non-linear tools wtll also be equippedwith the CAD system.
文摘A 3-Dimensional computer aided garment design (CAGD) system has been developed andimplemented on a high-performance workstation. We studied various approaches to the func-tional modelling of garment designs for the system. According to the characteristic data of a hu-man body, the models of human body and the garment are displayed on the screen, then we canmodify the garment with various styles and different sizes. The system can transform the 3-Dgarment to the 2-D pieces. The system has improved design efficiency. Various potential alterna-tives and improvement of the system have also been studied and explored.
文摘This paper presents a general method for 2D/3D transformation, which can be efficiently used in three dimensional computer aided garment design. The method utilizes a uniform triangular spring_mass based deformable model. 2D to 3D transformation and 3D to 2D transformation both can be implemented on the same model. A general and efficient collision detection method is also briefly discussed in this paper.
文摘目的经颅超声成像技术作为高效率、低成本且无创的诊断手段,已逐步应用于帕金森病患者认知功能障碍诊断。由于经颅超声图像信噪比低、成像质量差、目标组织复杂且相似度高,需要依赖专业医生手动检测。但是人工检测不仅费时费力,还可能因为操作者的主观因素影响,造成检测结果出现差异性。针对这一问题,提出了一种基于Swin Transformer和多尺度深度特征融合的YOLO-SF-TV(YOLO network based on Swin Transformer and multiscale deep feature fusion for third ventricle)模型用于经颅超声图像三脑室检测,以提高临床检测准确率,辅助医生进行早期诊断。方法YOLO-SF-TV模型在YOLOv8的基础上使用基于窗口注意力的Swin Transformer作为模型特征提取网络,并引入空间金字塔池化合模块SPP-FCM(spatial pyramid pooling fast incorporating CSPNet and multiple attention mechanisms)扩大网络感受野,并增强多尺度特征融合能力。在网络的多尺度特征融合部分结合深度可分离卷积和多头注意力机制,提出了PAFPN-DM(path aggregation and feature pyramid network with depthwise separable convolution)模块,并对主干特征输出层增加多头注意力机制,以提高网络对不同尺度特征图中全局和局部重要信息的理解能力。同时,将传统卷积替换为深度可分离卷积模块,通过对每个通道单独卷积提高网络对不同通道的敏感性,以保证模型准确度的同时降低训练参数和难度,增强模型的泛化能力。结果在本文收集的经颅超声三脑室图像数据及对应标签的数据集上进行实验,并与典型的目标检测模型对比。实验结果表明,本文提出的YOLO-SF-TV在经颅超声三脑室目标上的平均精确度均值(mean average precision,mAP)达到98.69%,相比于YOLOv8提升了2.12%,与其他典型模型相比检测精度达到最优。结论本文提出的YOLO-SF-TV模型在经颅超声图像三脑室检测问题上表现优秀,SPP-FCM模块和PAFPN-DM模块可以增强模型检测能力,提高模型泛化性和鲁棒性。同时,本文制作的数据集将有助于推动经颅超声三脑室图像检测问题的研究。
