期刊文献+
共找到3篇文章
< 1 >
每页显示 20 50 100
基于自适应学习率的深度信念网设计与应用 被引量:20
1
作者 乔俊飞 王功明 +2 位作者 李晓理 韩红桂 柴伟 《自动化学报》 EI CSCD 北大核心 2017年第8期1339-1349,共11页
针对深度信念网(Deep belief network,DBN)预训练耗时长的问题,提出了一种基于自适应学习率的DBN(Adaptive learning rate DBN,ALRDBN).ALRDBN将自适应学习率引入到对比差度(Contrastive divergence,CD)算法中,通过自动调整学习步长来提... 针对深度信念网(Deep belief network,DBN)预训练耗时长的问题,提出了一种基于自适应学习率的DBN(Adaptive learning rate DBN,ALRDBN).ALRDBN将自适应学习率引入到对比差度(Contrastive divergence,CD)算法中,通过自动调整学习步长来提高CD算法的收敛速度.然后设计基于自适应学习率的权值训练方法,通过网络性能分析给出学习率变化系数的范围.最后,通过一系列的实验对所设计的ALRDBN进行测试,仿真实验结果表明,ALRDBN的收敛速度得到了提高且预测精度也有所改善. 展开更多
关键词 深度信念网 自适应学习率 对比差度 收敛速度 性能分析
下载PDF
An improved detail enhancement algorithm based on difference curvature and contrast field
2
作者 刘祎 陈燕 桂志国 《Journal of Measurement Science and Instrumentation》 CAS CSCD 2016年第3期247-254,共8页
The gradient image is always sensitive to noise in image detail enhancement. To overcome this shortage, an improved detail enhancement algorithm based on difference curvature and contrast field is proposed. F... The gradient image is always sensitive to noise in image detail enhancement. To overcome this shortage, an improved detail enhancement algorithm based on difference curvature and contrast field is proposed. Firstly, the difference curvature is utilized to determine the amplification coefficient instead of the gradient. This new amplification function of the difference curvature takes more neighboring points into account, it is therefore not sensitive to noise. Secondly, the contrast field is nonlinearly amplified according to the new amplification coefficient. And then, with the enhanced contrast field, we construct the energy functional. Finally, the enhanced image is reconstructed by the variational method. Experimental results of standard testing image and industrial X-ray image show that the proposed algorithm can perform well on increasing contrast and sharpening edges of images while suppressing noise at the same time. 展开更多
关键词 image enhancement contrast field difference curvature variational enhancement scheme
下载PDF
Local information enhanced LBP
3
作者 张刚 苏光大 +1 位作者 陈健生 王晶 《Journal of Central South University》 SCIE EI CAS 2013年第11期3150-3155,共6页
Based on the observation that there exists multiple information in a pixel neighbor,such as distance sum and gray difference sum,local information enhanced LBP(local binary pattern)approach,i.e.LE-LBP,is presented.Geo... Based on the observation that there exists multiple information in a pixel neighbor,such as distance sum and gray difference sum,local information enhanced LBP(local binary pattern)approach,i.e.LE-LBP,is presented.Geometric information of the pixel neighborhood is used to compute minimum distance sum.Gray variation information is used to compute gray difference sum.Then,both the minimum distance sum and the gray difference sum are used to build a feature space.Feature spectrum of the image is computed on the feature space.Histogram computed from the feature spectrum is used to characterize the image.Compared with LBP,rotation invariant LBP,uniform LBP and LBP with local contrast,it is found that the feature spectrum image from LE-LBP contains more details,however,the feature vector is more discriminative.The retrieval precision of the system using LE-LBP is91.8%when recall is 10%for bus images. 展开更多
关键词 texture feature extraction LE-LBP minimum distance sum gray difference sum
下载PDF
上一页 1 下一页 到第
使用帮助 返回顶部