期刊名称：International Journal of Innovative Research in Computer and Communication Engineering
印刷版ISSN：2320-9798
电子版ISSN：2320-9801
出版年度：2019
卷号：7
期号：3
页码：2003-2007
DOI：10.15680/IJIRCCE.2019. 0703077
出版社：S&S Publications
摘要：The basic purpose behind the work displayed in this paper, is to separate the possibility of a Deep
Learning calculation to be particular, Convolutional neural systems (CNN) in picture characterization. A study of Deep
Learning, its methodologies, examination of structures, and calculations is introduced. The significance of (adequate)
training has been considered. Experimental results with generally utilized hyperspectral information show that
classifiers worked in this deep learning-based structure give focused execution. The advancement has shown
imperative execution in various vision assignments, for instance, image identification, question area and sementic
division. Specifically, late advances of deep learning systems pass on requesting that execution fine-grained image
classification which means to see subordinate-level classifications.
关键词：Convolutional Neural Networks; Image classification; Batch Normalization; Deep learning
