AIM: To investigate the diagnostic performance of multi-detector computed tomography (MDCT) in detecting biliary complications after orthotopic liver transplantation (OLT).
Abstract-A brain-computer interface （BCI） real- time system based on motor imagery translates the user＇s motor intention into a real-time control signal for peripheral equipments. A key problem to be solved for pra...Abstract-A brain-computer interface （BCI） real- time system based on motor imagery translates the user＇s motor intention into a real-time control signal for peripheral equipments. A key problem to be solved for practical applications is real-time data collection and processing. In this paper, a real-time BCI system is implemented on computer with electroencephalogram amplifier. In our implementation, the on-line voting method is adopted for feedback control strategy, and the voting results are used to control the cursor horizontal movement. Three subjects take part in the experiment. The results indicate that the best accuracy is 90%.展开更多
《计算》(Compuing Magazine of the CCF,CCCF)作为中国计算机学会(China Computer Federation,CCF)的会刊以及旗舰刊物,由工业和信息化部主管,哈尔滨工程大学主办,中国计算机学会指导。本刊专注于计算领域,旨在刊发本领域的研究成果及...《计算》(Compuing Magazine of the CCF,CCCF)作为中国计算机学会(China Computer Federation,CCF)的会刊以及旗舰刊物,由工业和信息化部主管,哈尔滨工程大学主办,中国计算机学会指导。本刊专注于计算领域,旨在刊发本领域的研究成果及应用,搭建学术交流平台,促进信息技术创新发展,助力读者和行业从业者洞悉计算技术前沿动态,推动科技创新与社会进步。展开更多
现有目标检测模型在边缘设备上部署时,其检测性能和推理速度的平衡有较大提升空间。针对此问题,本文基于YOLO(you can only look once)v8提出一种可部署到多类边缘设备上的目标检测模型。在模型的骨干网络部分,设计了EC2f(extended coar...现有目标检测模型在边缘设备上部署时,其检测性能和推理速度的平衡有较大提升空间。针对此问题,本文基于YOLO(you can only look once)v8提出一种可部署到多类边缘设备上的目标检测模型。在模型的骨干网络部分,设计了EC2f(extended coarse-to-fine)结构,在降低参数量和计算复杂度的同时降低数据读写量;在颈部网络部分,将颈部网络替换为YOLOv6-3.0版本的颈部网络,加速了模型推理,并将推理精度维持在较好水平;预测头网络部分设计了多尺度卷积检测头,进一步降低了模型的计算复杂度和参数度。设计了两个版本(n/s尺度)以适应不同的边缘设备。在X光数据集的实验表明,模型在推理精度上比同尺度的基准模型分别提升0.5/1.7百分点,推理速度上分别提升11.6%/11.2%。在其他数据集上的泛化性能测试表明,模型的推理速度提升了10%以上,精度降低控制在1.3%以内。实验证明,模型在推理精度和速度之间实现了良好的平衡。展开更多
基金Supported by National Natural Science Foundation of China,No.81201090,No.81371655Guangdong Natural Science Foundation,No.S2012010008367
文摘AIM: To investigate the diagnostic performance of multi-detector computed tomography (MDCT) in detecting biliary complications after orthotopic liver transplantation (OLT).
基金supported by the National Natural Science Foundation of China under Grant No. 60571019UESTC Youth Foundation under Grant No. L08010901JX0772 for support.
文摘Abstract-A brain-computer interface （BCI） real- time system based on motor imagery translates the user＇s motor intention into a real-time control signal for peripheral equipments. A key problem to be solved for practical applications is real-time data collection and processing. In this paper, a real-time BCI system is implemented on computer with electroencephalogram amplifier. In our implementation, the on-line voting method is adopted for feedback control strategy, and the voting results are used to control the cursor horizontal movement. Three subjects take part in the experiment. The results indicate that the best accuracy is 90%.
文摘《计算》(Compuing Magazine of the CCF,CCCF)作为中国计算机学会(China Computer Federation,CCF)的会刊以及旗舰刊物,由工业和信息化部主管,哈尔滨工程大学主办,中国计算机学会指导。本刊专注于计算领域,旨在刊发本领域的研究成果及应用,搭建学术交流平台,促进信息技术创新发展,助力读者和行业从业者洞悉计算技术前沿动态,推动科技创新与社会进步。
文摘现有目标检测模型在边缘设备上部署时,其检测性能和推理速度的平衡有较大提升空间。针对此问题,本文基于YOLO(you can only look once)v8提出一种可部署到多类边缘设备上的目标检测模型。在模型的骨干网络部分,设计了EC2f(extended coarse-to-fine)结构,在降低参数量和计算复杂度的同时降低数据读写量;在颈部网络部分,将颈部网络替换为YOLOv6-3.0版本的颈部网络,加速了模型推理,并将推理精度维持在较好水平;预测头网络部分设计了多尺度卷积检测头,进一步降低了模型的计算复杂度和参数度。设计了两个版本(n/s尺度)以适应不同的边缘设备。在X光数据集的实验表明,模型在推理精度上比同尺度的基准模型分别提升0.5/1.7百分点,推理速度上分别提升11.6%/11.2%。在其他数据集上的泛化性能测试表明,模型的推理速度提升了10%以上,精度降低控制在1.3%以内。实验证明,模型在推理精度和速度之间实现了良好的平衡。
