A method for enhancing the accuracy of image data sets for loss prevention applications in convolutional neural networks

摘要

method'Improved accuracy'data set'Image D'study'Oneneural networkconvolutionApplication for loss preventionThe techniques for improving the accuracy of the neural network formed for the application of damage prevention include: identifying the physical characteristics of the objects in the image scanning data (108), and abbreviating the marks of the image scanning data. In addition, neural network is used to detect the physical features in the scanned data of the suppressed tagged image according to the comparison of the recognition data based on physical characteristics and other recognition,In response to matching prediction, matching instructions are given and authentication signals are generated.Abbreviations: Figure 1