Jove
Visualize
Contact Us
JoVE
x logofacebook logolinkedin logoyoutube logo
ABOUT JoVE
OverviewLeadershipBlogJoVE Help Center
AUTHORS
Publishing ProcessEditorial BoardScope & PoliciesPeer ReviewFAQSubmit
LIBRARIANS
TestimonialsSubscriptionsAccessResourcesLibrary Advisory BoardFAQ
RESEARCH
JoVE JournalMethods CollectionsJoVE Encyclopedia of ExperimentsArchive
EDUCATION
JoVE CoreJoVE BusinessJoVE Science EducationJoVE Lab ManualFaculty Resource CenterFaculty Site
Terms & Conditions of Use
Privacy Policy
Policies

Related Experiment Video

Updated: May 18, 2026

End-To-End Deep Neural Network for Salient Object Detection in Complex Environments
03:31

End-To-End Deep Neural Network for Salient Object Detection in Complex Environments

Published on: December 15, 2023

A simultaneous dual watermarking scheme for deep learning models.

Dehui Wang1, Yingqian Zhang2, Shuang Zhou3

  • 1School of Mathematical Sciences, Beihang University, Beijing, 100083, China.

Neural Networks : the Official Journal of the International Neural Network Society
|May 16, 2026
PubMed
Summary
This summary is machine-generated.

Related Concept Videos

You might also read

Related Articles

Articles linked to this work by shared authors, journal, and citation graph.

Sort by
Same author

A novel surgical T staging system for recurrent nasopharyngeal carcinoma: advancing prognostic accuracy and clinical applicability.

Journal of neurosurgery·2026
Same author

Efficacy Analysis of Simultaneous Bariatric Surgery and Radical Thyroidectomy in Patients with Obesity and Low-Risk Papillary Thyroid Cancer: A Pilot Feasibility and Safety Study.

Obesity surgery·2026
Same author

Replenishing interfacial lubrication: A cartilage-inspired polyzwitterionic dressing for spatiotemporal healing of friction-prone diabetic wound.

Biomaterials·2026
Same author

Endoscopic endonasal surgery and radiotherapy in skull base chordomas: a 20-year retrospective analysis of long-term outcomes and prognostic factors.

Acta oto-laryngologica·2026
Same author

Effects of Combined Diet and Physical Activity on Gestational Weight Gain in Low-Risk Pregnant Women Based on the TIDieR Checklist: A Systematic Review and Meta-Analysis.

Healthcare (Basel, Switzerland)·2026
Same author

Effects of gestational weight gain based on weight-recommended guidelines (2022) in China on maternal and neonatal outcomes: a retrospective cohort study.

BMC public health·2026
Same journal

Anchor-based disentanglement framework for incremental multi-view clustering.

Neural networks : the official journal of the International Neural Network Society·2026
Same journal

Complex-valued amplitude-phase interference modeling for adversarially robust classification.

Neural networks : the official journal of the International Neural Network Society·2026
Same journal

TraNce: Type-aware hypergraph neural network with biological mediators for drug repositioning.

Neural networks : the official journal of the International Neural Network Society·2026
Same journal

Decentralized ADMM for factorization-based Low-rank matrix estimation.

Neural networks : the official journal of the International Neural Network Society·2026
Same journal

Memristive neuromorphic circuit design inspired by the neural mechanisms of conditioned fear.

Neural networks : the official journal of the International Neural Network Society·2026
Same journal

Q-learning based asynchronous Boolean control networks stabilization with data loss.

Neural networks : the official journal of the International Neural Network Society·2026
See all related articles

This study introduces a dual watermarking scheme to protect intellectual property for both deep learning model developers and users. The novel method ensures robust protection against various attacks without compromising model performance.

Area of Science:

  • Computer Science
  • Artificial Intelligence
  • Machine Learning

Background:

  • Deep learning models (DLMs) require robust intellectual property (IP) protection.
  • Existing watermarking methods often protect only the seller or buyer, not both simultaneously.
  • Dual ownership and distribution of DLMs necessitate a comprehensive IP protection strategy.

Purpose of the Study:

  • To propose a simultaneous dual watermarking scheme for DLMs.
  • To enable simultaneous IP protection for both sellers (developers) and buyers (users).
  • To ensure traceability and ownership verification for DLMs.

Main Methods:

  • Developed a dual watermarking scheme using two distinct trigger sets.
  • Integrated trigger sets and original datasets into a unified training set.
Keywords:
ChaosDeep learning modelDual watermarkingIntellectual property protectionOwnership verification

Related Experiment Videos

Last Updated: May 18, 2026

End-To-End Deep Neural Network for Salient Object Detection in Complex Environments
03:31

End-To-End Deep Neural Network for Salient Object Detection in Complex Environments

Published on: December 15, 2023

  • Utilized chaotic sequences for trigger set annotation, ensuring feature orthogonality.
  • Leveraged the perpendicular relationship between dataset features to maintain model performance.

    • Main Results:

    • The dual watermarking scheme effectively protects DLM IP for both buyers and sellers.
    • The method demonstrates superior effectiveness, fidelity, integrity, and robustness against attacks.
    • Statistical analysis confirmed no adverse impact on DLM decision boundaries or significant statistical characteristics.
    • The scheme proved robust against fine-tuning, overwriting, and fraudulent ownership claims.

    Conclusions:

    • The proposed simultaneous dual watermarking scheme offers a comprehensive solution for DLM IP protection.
    • This approach addresses the limitations of single watermark formats in protecting distributed DLMs.
    • The method provides enhanced security and verifiable ownership for deep learning models.