A Novel Visualization Malware Detection Method based on Spp-Net
Peng Zhang, Bowen Sun, Ruotong Ma, Ang Li · 2019
In recent years, the threat of malware in Internet is extremely serious, malware detection becomes one of the most challenging task in cyberspace. In this paper, we propose a novel malware detection method using code image and nerual network. We enhanced traditional gray code image with visible string and PE file structure to a RGB image, and then we put the image into the top 13 layer of VGG16, finally we use Spp-Net to adapt the feature size. In order to make label reliable, we use various Virustotal result to get a voting result, and using the algorithm to train and predict the malware. We test the proposed method in real malware datasets, experimental results show that our method has a better performance.