传统的Bag of Words模型检索方法并不具备局部特征间的空间关系,因此影响检索性能.本文提出了基于分级显著信息的空间编码方法.通过分层次的提取显著区域并对每个显著区域内的特征点进行空间编码.目的是探索特征间的空间关系,并根据分...
详细信息
传统的Bag of Words模型检索方法并不具备局部特征间的空间关系,因此影响检索性能.本文提出了基于分级显著信息的空间编码方法.通过分层次的提取显著区域并对每个显著区域内的特征点进行空间编码.目的是探索特征间的空间关系,并根据分级显著信息提高特征间的相关性.在几何验证过程中,本文通过任意三点间的角度编码和位移编码构成的空间编码方法完成图像对之间的空间关系匹配,同时根据图像各个区域间的显著程度赋予该区域空间关系匹配得分相应权重,得到最终的几何得分,重新排列检索结果.实验结果表明本文提出的方法既改善了最终检索结果的精确度又降低了几何验证阶段的计算时间.
In many informational fields,such as biological,environmental,medical,etc.,lots of sets of data are created every day.A flexible system biology tool enables to analysis the biology metabolize network in question with ...
详细信息
In many informational fields,such as biological,environmental,medical,etc.,lots of sets of data are created every day.A flexible system biology tool enables to analysis the biology metabolize network in question with efficiency and *** work gave a hybrid system based on the deep clustering algorithm,computing the co-regulated genes,also called regulons or co-regulation *** experiment results,based on the hypergeometric distribution function scores,showed that the model-based deep clustering (MBDC),was more efficient and accurate than other network functions as ISA,BIMAX,XMOTIFS,QUBIC and *** further study showed that MBDC was not only available in cancer analysis,but also in some bacteria study such as Eschedchia coli.
暂无评论