English (United Kingdom)

https://curated-unify.zendy.io/wp-json/zendy-region/v1/featured_content/oa?rat=en

https://curated-unify.zendy.io/wp-json/zendy-region/v1/highlighted_journal/

Global: Zendy Tools annual

Presents the keyphrase highlighting as premium feature

Keyphrase Highlighting

Presents the summarisation as premium feature

Summarisation

Insights

Presents the pdf analysis as premium feature

PDF Analysis

Presents the zaia usage as premium feature

ZAIA

Global: Zendy Tools monthly

Global: Zendy Plus monthly

Presents the access of premium content as premium feature

Premium Content

Global: Zendy Plus annual

Global: Zendy Free Plan

Intellectual property (IP) protection for Deep Neural Networks (DNNs) hasraised serious concerns in recent years. Most existing works embed watermarksin the DNN model for IP protection, which need to modify the model and lack ofinterpretability. In this paper, for the first time, we propose aninterpretable intellectual property protection method for DNN based onexplainable artificial intelligence. Compared with existing works, the proposedmethod does not modify the DNN model, and the decision of the ownershipverification is interpretable. We extract the intrinsic features of the DNNmodel by using Deep Taylor Decomposition. Since the intrinsic feature iscomposed of unique interpretation of the model's decision, the intrinsicfeature can be regarded as fingerprint of the model. If the fingerprint of asuspected model is the same as the original model, the suspected model isconsidered as a pirated model. Experimental results demonstrate that thefingerprints can be successfully used to verify the ownership of the model andthe test accuracy of the model is not affected. Furthermore, the proposedmethod is robust to fine-tuning attack, pruning attack, watermark overwritingattack, and adaptive attack.

InFIP: An Explainable DNN Intellectual Property Protection Method based on Intrinsic Features