一句话：提出基于框架的划分与隔离方法，加固数据处理软件以减轻框架漏洞利用的影响。
数据处理型软件，尤其是机器学习应用，高度依赖 TensorFlow、OpenCV 等标准框架与库。随着这些框架广泛使用，其漏洞被利用已成为关键安全问题。软件隔离本可降低漏洞利用影响，但现有方法难以分析复杂的程序依赖，或带来过高开销，因而实践中效果有限。FreePart 提出基于框架的划分与隔离方案，旨在对数据处理软件进行加固：围绕框架边界对程序进行划分并实施隔离，以限制框架漏洞被利用后的影响，同时规避复杂依赖分析与高开销问题。摘要未给出具体实验数据、基准或性能指标，评估细节有待论文正文补充。
Data processing oriented software, especially machine learning applications, are heavily dependent on standard frameworks/libraries such as TensorFlow and OpenCV. As those frameworks have gained significant popularity, the exploitation of vulnerabilities in the frameworks has become a critical security concern. While software isolation can minimize the impact of exploitation, existing approaches suffer from difficulty analyzing complex program dependencies or excessive overhead, making them ineffective in practice.
@inproceedings{paperbot10129,
title = {FreePart: Hardening Data Processing Software via Framework-based Partitioning and Isolation},
author = {Ali Ahad and Gang Wang and Chung Hwan Kim and Suman Jana and Zhiqiang Lin and Yonghwi Kwon},
booktitle = {ASPLOS},
year = {2023},
doi = {10.1145/3623278.3624760}
}