Advances in Binary Code Similarity Detection and Security

The field of binary code similarity detection and security is rapidly evolving, with a focus on improving the robustness of models against adversarial attacks and developing more efficient and effective methods for analyzing and rewriting binary code. Recent research has leveraged techniques such as explainers and semantic graphs to enhance the accuracy and resilience of binary code similarity analysis, while also exploring new approaches to taint tracking and vulnerability detection. Furthermore, there is a growing interest in applying AI-based methods to software vulnerability detection, with graph-based models emerging as a promising approach. Noteworthy papers in this area include:

  • A novel optimization for adversarial attacks against binary code similarity detection models, which achieves higher attack success rates and improved efficiency.
  • ORCAS, an obfuscation-resilient binary code similarity analysis model, which significantly outperforms existing approaches on the BinKit dataset.
  • A systematic literature review on AI-based software vulnerability detection, which provides a comprehensive taxonomy of techniques and identifies key limitations and emerging opportunities.

Sources

Explainer-guided Targeted Adversarial Attacks against Binary Code Similarity Detection Models

Obfuscation-Resilient Binary Code Similarity Analysis using Dominance Enhanced Semantic Graph

Scalable Language Agnostic Taint Tracking using Explicit Data Dependencies

A Systematic Literature Review on Continuous Integration and Deployment (CI/CD) for Secure Cloud Computing

Exploiting Control-flow Enforcement Technology for Sound and Precise Static Binary Disassembly

Empirical Quantification of Spurious Correlations in Malware Detection

Mapping NVD Records to Their VFCs: How Hard is it?

AI-Based Software Vulnerability Detection: A Systematic Literature Review

Built with on top of