Identification of Malicious Code Variants using Spp-Net Model and Color Images

With the rapid growth of the internet, security breaches have also increased employing malicious code attacks. Current methods for the detection of these codes require much improvement over the use of the same size dataset images and poor features of greyscale images. This paper proposed a method using the SPP-net model which can accept images of various sizes as input and also color images which provide many features for the detection of variants. Since the addition of a sublayer is required frequently, deep learning concept is incorporated. Also, they improve the detection of malicious variants too. Experimentation is done using CNN for the classification and SPP- net for various size images. Thus, the CNN architecture used in our proposed work is VGG16 which can deal with large scale recognition.