/he-man-concrete

HE-MAN – Homomorphically Encrypted MAchine learning with oNnx models and Concrete

Primary LanguageRustApache License 2.0Apache-2.0

he-man-concrete

Architecture

Usage

To generate a set of keys into the folder <SECRET_KEY_PATH>:

cargo run -- keygen <SECRET_KEY_PATH>

To encrypt the plaintext stored in file <PLAINTEXT_INPUT_PATH> using the set of keys stored in folder <SECRET_KEY_PATH> and write the resulting ciphertext into the file <CIPHERTEXT_OUTPUT_PATH> (not implemented):

cargo run -- encrypt <SECRET_KEY_PATH> <PLAINTEXT_INPUT_PATH> <CIPHERTEXT_OUTPUT_PATH>

To evaluate the model stored in file <MODEL_PATH> for the ciphertext input stored in file <CIPHERTEXT_INPUT_PATH> using the evaluation key stored in folder <EVALUATION_KEY_PATH> and write the encrypted model output into the file <CIPHERTEXT_OUTPUT_PATH> (not implemented):

cargo run -- inference <EVALUATION_KEY_PATH> <MODEL_PATH> <CIPHERTEXT_INPUT_PATH> <CIPHERTEXT_OUTPUT_PATH>

To decrypt the ciphertext stored in file <CIPHERTEXT_INPUT_PATH> using the set of keys stored in folder <SECRET_KEY_PATH> and write the resulting plaintext into the file <PLAINTEXT_OUTPUT_PATH> (not implemented):

cargo run -- decrypt <SECRET_KEY_PATH> <CIPHERTEXT_INPUT_PATH> <PLAINTEXT_OUTPUT_PATH>

To show how the program or a specific command is used:

cargo run -- --help
cargo run -- <COMMAND> --help

Dev setup in VSCode

use this command to clone repo and create a dev container, then use command above to run

Remote-Containers: Clone Repository in Container Volume...

Evaluation

Setup

  • Install Rust and Cargo: curl https://sh.rustup.rs -sSf | sh
  • Install Python (3.9)
  • Install Python dependencies: pip install argparse pathlib numpy tqdm
  • cd he-man-concrete
  • Build he-man-concrete: RUSTFLAGS="-C target-cpu=native" cargo build --release

Run evaluations

  • cd evaluation
  • Run evaluation: python benchmark.py -m cryptonets (Attention: This evaluation may take several minutes to hours!)
  • Parameters:
    • -m: model [cryptonets, lenet5 (default), mobilefacenets]
    • -s: start-index of samples, default: 0
    • -n: number of samples, default: 10

License

This software uses several dependencies that have different licenses. It is the responsibility of the user of this software to ensure they have the proper rights to the underlying dependencies.

citing HE-MAN

@inproceedings{10.1145/3589883.3589889,
  author = {Nocker, Martin and Drexel, David and Rader, Michael and Montuoro, Alessio and Sch\"{o}ttle, Pascal},
  title = {HE-MAN – Homomorphically Encrypted MAchine Learning with ONnx Models},
  year = {2023},
  isbn = {9781450398329},
  publisher = {Association for Computing Machinery},
  address = {New York, NY, USA},
  url = {https://doi.org/10.1145/3589883.3589889},
  doi = {10.1145/3589883.3589889},
  booktitle = {Proceedings of the 2023 8th International Conference on Machine Learning Technologies},
  pages = {35–45},
  numpages = {11},
  keywords = {Machine Learning as a Service, Homomorphic Encryption, Secure and Privacy-Preserving Machine Learning},
  location = {Stockholm, Sweden},
  series = {ICMLT '23}
}