Malware attacks are increasing exponentially with usage of the internet. The first step towards safeguarding from
malware attacks is by distinguishing malware files from benign ones and classifying them to known classes. Classification of
malware is helpful for the analyst as it helps them to get a better insight into the functioning of the malware. This paper
proposes a classification system for the malware variants into there families using Convolutional Neural Network along with
Spatial Pyramid Pooling layer. This system involves the visualization of malware binary file and uses the texture based similarity
in the images of same families for classification. Convolutional Neural Network along with Spatial Pyramid Pooling layer allows
to use multi scale images which improved the classification accuracy.
