Extending Null Embedding for Deep Neural Network (DNN) Watermarking

Kaan Altınay, Devriş İşler, Devriş İşler, Zekeriya Erkin

2025

Abstract

The rise of Machine Learning (ML) has opened new business opportunities, particularly through Machine Learning as a Service (MLaaS), where costly models like Deep Neural Networks (DNNs) can be outsourced. However, this also raises concerns about model piracy. To protect against unauthorized use, watermarking techniques have been developed. One such method, null embedding by Li et al., disables the model if pirated but reduces classification accuracy. This paper proposes modifications to the null-embedding technique that reduce this impact and keep the classification accuracy close to that of a non-watermarked model.

Paper Citation

in Harvard Style

Altınay K., İşler D. and Erkin Z. (2025). Extending Null Embedding for Deep Neural Network (DNN) Watermarking. In Proceedings of the 22nd International Conference on Security and Cryptography - Volume 1: SECRYPT; ISBN 978-989-758-760-3, SciTePress, pages 771-776. DOI: 10.5220/0013641200003979

in Bibtex Style

@conference{secrypt25,
author={Kaan Altınay and Devriş İşler and Zekeriya Erkin},
title={Extending Null Embedding for Deep Neural Network (DNN) Watermarking},
booktitle={Proceedings of the 22nd International Conference on Security and Cryptography - Volume 1: SECRYPT},
year={2025},
pages={771-776},
publisher={SciTePress},
organization={INSTICC},
doi={10.5220/0013641200003979},
isbn={978-989-758-760-3},
}

in EndNote Style

TY - CONF

JO - Proceedings of the 22nd International Conference on Security and Cryptography - Volume 1: SECRYPT
TI - Extending Null Embedding for Deep Neural Network (DNN) Watermarking
SN - 978-989-758-760-3
AU - Altınay K.
AU - İşler D.
AU - Erkin Z.
PY - 2025
SP - 771
EP - 776
DO - 10.5220/0013641200003979
PB - SciTePress