Four-Part Grayscale Images for Image-Based Malware Detection

This dataset contains 16,019 malware and benign samples represented as grayscale images, including 11,971 malware samples from 35 families and 4,048 benign files. Each original executable was divided into four distinct sections (text, .data, .rsrc, and header) and converted into grayscale images. It is organised into training, validation, and testing subsets with set splits for both malware and benign classes: Training (8,291 malware; 2,833 benign), Validation (1,833 malware; 607 benign) and Testing (1,847 malware; 608 benign).