EUVIMEDOpdag EUVIMED
Open access · CC BY 4.0

Unnoticeable Hybrid Watermarking for Deep Neural Network Authentication Using Auxiliary Hidden Layers

Rodrigo Eduardo Arevalo-Ancona, Manuel Cedillo-Hernandez

Machine Learning and Knowledge Extraction · 2026-05-22 · 8 (6) · 140

DOI: 10.3390/make8060140 ↗

At a glance

Journal article in Machine Learning and Knowledge Extraction, published online on 2026-05-22. Written by Rodrigo Eduardo Arevalo-Ancona and 1 co-authors. 33 references. The full text is freely available (open access, CC BY 4.0). Funded by: Secretaría de Ciencia, Humanidades, Tecnología e Innovación (SECIHTI) of Mexico, Secretaría de Investigación y Posgrado of the Instituto Politécnico Nacional.

AbstractBibliografiske oplysningerCiteRelated studies

Abstract (original language)

The authentication and protection of deep neural network models have become challenging due to their widespread distribution and reuse, making them vulnerable to unauthorized access. This paper addresses the need for ownership verification by proposing a hybrid neural network watermarking method for secure model authentication. The approach combines a steganographic watermark embedded into stable model weights with a user code for watermark recovery encoded in auxiliary hidden layers. Stable parameters are identified through a reduced training to estimate gradient variations for the watermark insertion with minimal impact on model performance. Additionally, two auxiliary layers are introduced, to store in the first layer the metadata indices from the selected weights where the watermark was embedded and in the second layer the user code, supporting secure identification and verification. Experimental evaluations demonstrate that the proposed method remains robust under different model optimization attacks, including pruning, fine-tuning, additive noise injection, and parameter overwriting, while preserving model performance. The proposed framework achieves a BER = 0 under several moderate attack scenarios across different neural network models, whereas more aggressive optimizations degrade the watermark recovery performance. These results indicate that the proposed framework provides an effective solution for neural network ownership protection while maintaining the model performance.

Funding

  • Secretaría de Ciencia, Humanidades, Tecnología e Innovación (SECIHTI) of Mexico
  • Secretaría de Investigación y Posgrado of the Instituto Politécnico Nacional

Cite

APA

Arevalo-Ancona, R. E., & Cedillo-Hernandez, M. (2026). Unnoticeable Hybrid Watermarking for Deep Neural Network Authentication Using Auxiliary Hidden Layers. Machine Learning and Knowledge Extraction, 8(6), 140. https://doi.org/10.3390/make8060140

Vancouver

Arevalo-Ancona RE, Cedillo-Hernandez M. Unnoticeable Hybrid Watermarking for Deep Neural Network Authentication Using Auxiliary Hidden Layers. Machine Learning and Knowledge Extraction. 2026;8(6):140. doi:10.3390/make8060140

BibTeX
@article{ArevaloAncona2026,
  author = {Arevalo-Ancona, Rodrigo Eduardo and Cedillo-Hernandez, Manuel},
  title = {Unnoticeable Hybrid Watermarking for Deep Neural Network Authentication Using Auxiliary Hidden Layers},
  journal = {Machine Learning and Knowledge Extraction},
  year = {2026},
  volume = {8},
  number = {6},
  pages = {140},
  doi = {10.3390/make8060140}
}
Start for free →