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/

Zendy Plus

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

Zendy Tools

Zendy Open

Recently, machine learning (ML) has introduced advanced solutions to manydomains. Since ML models provide business advantage to model owners, protectingintellectual property of ML models has emerged as an important consideration.Confidentiality of ML models can be protected by exposing them to clients onlyvia prediction APIs. However, model extraction attacks can steal thefunctionality of ML models using the information leaked to clients through theresults returned via the API. In this work, we question whether modelextraction is a serious threat to complex, real-life ML models. We evaluate thecurrent state-of-the-art model extraction attack (Knockoff nets) againstcomplex models. We reproduce and confirm the results in the original paper. Butwe also show that the performance of this attack can be limited by severalfactors, including ML model architecture and the granularity of API response.Furthermore, we introduce a defense based on distinguishing queries used forKnockoff nets from benign queries. Despite the limitations of the Knockoffnets, we show that a more realistic adversary can effectively steal complex MLmodels and evade known defenses.

Extraction of Complex DNN Models: Real Threat or Boogeyman?