MalSort: Lightweight and efficient image-based malware classification using masked self-supervised framework with Swin Transformer

Fangwei Wang, Xipeng Shi, Fang Yang, Ruixin Song, Qingru Li, Zhiyuan Tan, Changguang Wang

Abstract: Highlights•An asymmetric structure extracts features to handle labeled samples and feature loss.•The fine-tune incorporates class re-balancing to handle imbalance and overfitting.•A designed lightweight module improves convergence and decreases resource dependence.

External IDs:doi:10.1016/j.jisa.2024.103784