Using binary malware data, which has been processed into visual input, we can detect varaiant malware at very high accuarcy. More work has been conducted on this by Lakshmanan Nataraj:
https://vizsec.org/files/2011/Nataraj.pdf
https://vision.ece.ucsb.edu/sites/default/files/publications/nataraj_vizsec_2011_paper.pdf
This repo will be dedicated to improving the accuracy achieved by previous researchers by using state of the art neural networks architectures.