Subjects: Computer Science >> Integration Theory of Computer Science submitted time 2022-05-18 Cooperative journals: 《计算机应用研究》
Abstract： Bone age assessment is a common method to detect endocrine and growth abnormalities in children. But in deep learning methods, low-quality hand X-ray images reduce the final evaluation accuracy. To solve this problem, this paper proposed an alignment network that increases the area of interest in hand X-ray images. This network uses the Swin Transformer structure as the backbone network to learn image hand similarity and obtain affine coefficients and does not require large-scale hand annotation during the training process. In the bone age assessment network, for the improvement of efficient channel attention and spatial attention mechanism. This paper proposed dual-pool efficient channel attention and asymmetric convolution spatial attention method and combines these two methods in the form of dual attention and Xception network proposes DA-Xception. This paper tested the RSNA dataset, which achieved a mean absolute error of 5.37 months for this bone age assessment method. Compared with other deep learning methods, this method can fully extract features and optimize the evaluation results.
Subjects: Computer Science >> Integration Theory of Computer Science submitted time 2018-09-12 Cooperative journals: 《计算机应用研究》
Abstract： Aiming at improving the working efficiency and resource utilization of the polling control system, a limited (K=2) polling control system for the site dormant activity state is proposed. The mathematical model of the system is studied by using the probabilistic parent function and the embedded Markov chain method. The simulation results show that the theoretical value is approximately equal to the experimental value, which indicates that the analysis method is correct and reasonable. Based on the limited (K=2) policy, the system serves the server according to the state of the dormant activity of the site. The qualified (K=2) service policy is used to ensure that the fairness of the system is not impaired, and distinguishing the dormant activity state of the site avoids the service to the dormant site without information packets. Compared with the existing service strategy, the system performance is significantly improved.
Subjects: Computer Science >> Integration Theory of Computer Science submitted time 2018-08-13 Cooperative journals: 《计算机应用研究》
Abstract： In order to improve the efficiency and resource utilization of polling control system and ensure system fairness is not impaired, a limited (K=2) polling control system was proposed to distinguish the busy and idle status of the site. The mathematical model of the system was studied by using probability generating function and embedded Markov chain. The simulation results show that: the theoretical values and the experimental values are approximately equal, indicating that the analysis method is correct and reasonable. The system is based on a limited (K=2) polling service strategy, according to the different status of the site, only dynamic busy sites with information groups are served. The limited (K=2) service strategy ensures the system fairness, and distinguishing busy and idle status of site avoids the query service for the idle site without information groups, thereby improving system efficiency and resource utilization. Compared with existing service strategy, the system performance is obviously improved.
Subjects: Computer Science >> Integration Theory of Computer Science submitted time 2018-05-20 Cooperative journals: 《计算机应用研究》
Abstract： In embedded application fields, the operating system presentes the more stern reality demand such as reconstruction, expansion, transplantation, interaction, security, and efficient. In combination with the above features of embedded system and inherent characteristics of Forth systems, using Forth virtual machine technology, this paper explored the key technologies of embedded operating system based on Forth virtual machine architecture, and proposed an embedded multitask operating system scheduling algorithm based on Forth virtual machine architecture, which had better extension and transplantation characteristics. The algorithm adopted a cooperative multi-task scheduling mechanism with Forth virtual machine instruction synchronization, shortened the task switching time, and simplified context switching operations to simply saving the data stack pointer. The experiment results demonstrate that the multitask scheduling algorithm based on Forth virtual machine architecture develops the inherent characteristics of Forth system, presents higher efficiency for specific applications, especially benefits for embedded environments with limited resources.
Subjects: Computer Science >> Integration Theory of Computer Science submitted time 2018-05-20 Cooperative journals: 《计算机应用研究》
Abstract： Faced with more and more complex embedded application requirements and urgent problems of the current embedded operating system research areas, such as reconstruction, transplantation, maintenance, trusted, multi-core, many-core and so on, this paper proposed an embedded multitask operating system architecture, which presents better scalability, transplantation, streamlined and efficient by exploring the key technologies of embedded operating system based on Forth virtual machine technology. The architecture implemented code sharing and multitask managing, by using the methods of classification storage mapping, Forth vector definition and user variable separation. The experiment results demonstrate that the embedded operating system architecture based on Forth virtual machine takes advantage of the inherent characteristics of the system, reduces the resources consumption, and improves system flexibility and operational efficiency.
Hits
Hits
Downloads
Comment