Proj CDeepFuzz Paper Reading: RapidFuzz: Accelerating fuzzing via Generative Adversarial Networks

Abstract

本文: RapidFuzz
Task: use GAN to fuzz program with highly-structured inputs
Method: use GAN to capture the data structure, then locate the hot-points generated by GAN,
BaseTool: AFL

实验：
Competitor: AFL
效果:

  1. improved coverage, For tiff2pdf and tiffdump, coverage increase exceeds over 20%.
  2. RapidFuzz achieves the same coverage with less time
  3. AFL absorbs 21% of generated seed files in tiff2pdf with an average absorption rate around 15% in other programs
posted @ 2023-08-06 20:35  雪溯  阅读(23)  评论(0)  编辑  收藏  举报