/isadetect

ML-based ISA detection (architecture and endianness of binary code/sequences)

Primary LanguagePythonMIT LicenseMIT

Project

ISAdetect - "ML-based ISA detection (architecture and endianness of binary code/sequences)"

Live demo at https://isadetect.com/

Referencing

If/when using code and/or data from this project (direct, derivative works), please don't forget to reference the following:

@inproceedings{kairajarvi2020isadetect,
author={Kairaj\"arvi, Sami and Costin, Andrei and H\"am\"al\"ainen, Timo},
title={{ISAdetect: Usable Automated Detection of CPU Architecture and Endianness for Executable Binary Files and Object Code}},
booktitle={Proceedings of the Tenth ACM Conference on Data and Application Security and Privacy},
year={2020},
url="https://doi.org/10.1145/3374664.3375742"
}

Directory structure

ml: Machine Learning related scripts: training, generating ML model.

api: Exposing the ML classification as RESTful API via Swagger/OpenAPI.

dataset_gen: Set of tools and scripts to gather and compile the datasets.

plugins: Plugins that call the RESTful API from SotA tools such as radare2 and IDAPro.

Dataset

Dataset related to this project can be found at https://etsin.fairdata.fi/dataset/9f6203f5-2360-426f-b9df-052f3f936ed2 (http://urn.fi/urn:nbn:fi:att:693a3e3a-976a-4eac-8c3d-a4a62619f8b1)

Questions and comments

Please feel free to submit an issue on Github for any questions or concerns.