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

The prosperity of deep neural networks (DNNs) is largely benefited fromopen-source datasets, based on which users can evaluate and improve theirmethods. In this paper, we revisit backdoor-based dataset ownershipverification (DOV), which is currently the only feasible approach to protectthe copyright of open-source datasets. We reveal that these methods arefundamentally harmful given that they could introduce maliciousmisclassification behaviors to watermarked DNNs by the adversaries. In thispaper, we design DOV from another perspective by making watermarked models(trained on the protected dataset) correctly classify some `hard' samples thatwill be misclassified by the benign model. Our method is inspired by thegeneralization property of DNNs, where we find a \emph{hardly-generalizeddomain} for the original dataset (as its \emph{domain watermark}). It can beeasily learned with the protected dataset containing modified samples.Specifically, we formulate the domain generation as a bi-level optimization andpropose to optimize a set of visually-indistinguishable clean-label modifieddata with similar effects to domain-watermarked samples from thehardly-generalized domain to ensure watermark stealthiness. We also design ahypothesis-test-guided ownership verification via our domain watermark andprovide the theoretical analyses of our method. Extensive experiments on threebenchmark datasets are conducted, which verify the effectiveness of our methodand its resistance to potential adaptive methods. The code for reproducing mainexperiments is available at\url{https://github.com/JunfengGo/Domain-Watermark}.

Domain Watermark: Effective and Harmless Dataset Copyright Protection is Closed at Hand