中国科学院机构知识库网格系统: 检索

浏览/检索结果: 共3条，第1-3条

帮助

条数/页：排序方式：
	Fusion of heterogeneous attention mechanisms in multi-view convolutional neural network for text classification 期刊论文 OAI收割 INFORMATION SCIENCES, 2021, 卷号: 548, 页码: 295-312 作者: Liang, Yunji; Li, Huihui; Guo, Bin; Yu, Zhiwen; Zheng, Xiaolong \| 收藏 \| 浏览/下载：36/0 \| 提交时间：2021/03/08 View attention Spatial attention Multi-view representation Series and parallel connection Conventional neural network Text classification
	Wavelet packet and neural network basis medical image compression (EI CONFERENCE) 会议论文 OAI收割 4th International Conference on Photonics and Imaging in Biology and Medicine, September 3, 2005 - September 6, 2005, Tianjin, China Zhao X.; Wei J.; Zhai L. 收藏 \| 浏览/下载：21/0 \| 提交时间：2013/03/25 It is difficult to get high compression ratio and good reconstructed image by conventional methods we give a new method of compression on medical image. It is to decompose and reconstruct the medical image by wavelet packet. Before the construction the image use neural network in place of other coding method to code the coefficients in the wavelet packet domain. By using the Kohonen's neural network algorithm not only for its vector quantization feature but also for its topological property. This property allows an increase of about 80% for the compression rate. Compared to the JPEG standard this compression scheme shows better performances (in terms of PSNR) for compression rates higher than 30. This method can get big compression ratio and perfect PSNR. Results show that the image can be compressed greatly and the original image can be recovered well. In addition the approach can be realized easily by hardware.
	An improved adaptive neural network method for control system (EI CONFERENCE) 会议论文 OAI收割 2006 International Conference on Machine Learning and Cybernetics, August 13, 2006 - August 16, 2006, Dalian, China Wang L.-M.; Xie M.-J.; Wu D.-Y. 收藏 \| 浏览/下载：18/0 \| 提交时间：2013/03/25 Classical methods for designing a controller depend on the accuracy of system model. However plant's models and other parts in a physical system can not accurately represent all possible dynamics. Thus the controller designed is usually not the optimal one. In this article a new simple adaptive control method which combines the classical frequency domain method with the neural network theory is proposed. Firstly we can obtain a controller using classical method. Secondly we use the coefficients in digitized controller equation as the initial values of an Adaline network. Finally LMS learning rules is used to adjust the weights adaptively. Experimental results show that this method is very effective in improving the performance of conventional controller. 2006 IEEE.