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 Plus annual

Presents the access of premium content as premium feature

Premium Content

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 Plus monthly

Global: Zendy Tools annual

Global: Zendy Tools monthly

Global: Zendy Free Plan

Deep neural networks have had enormous impact on various domains of computerscience, considerably outperforming previous state of the art machine learningtechniques. To achieve this performance, neural networks need large quantitiesof data and huge computational resources, which heavily increases theirconstruction costs. The increased cost of building a good deep neural networkmodel gives rise to a need for protecting this investment from potentialcopyright infringements. Legitimate owners of a machine learning model want tobe able to reliably track and detect a malicious adversary that tries to stealthe intellectual property related to the model. Recently, this problem wastackled by introducing in deep neural networks the concept of watermarking,which allows a legitimate owner to embed some secret information(watermark) ina given model. The watermark allows the legitimate owner to detect copyrightinfringements of his model. This paper focuses on verifying the robustness andreliability of state-of- the-art deep neural network watermarking schemes. Weshow that, a malicious adversary, even in scenarios where the watermark isdifficult to remove, can still evade the verification by the legitimate owners,thus avoiding the detection of model theft.

Have You Stolen My Model? Evasion Attacks Against Deep Neural Network Watermarking Techniques