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

Recent studies shows that the majority of existing deep steganalysis modelshave a large amount of redundancy, which leads to a huge waste of storage andcomputing resources. The existing model compression method cannot flexiblycompress the convolutional layer in residual shortcut block so that asatisfactory shrinking rate cannot be obtained. In this paper, we proposeSTD-NET, an unsupervised deep-learning architecture search approach viahierarchical tensor decomposition for image steganalysis. Our proposed strategywill not be restricted by various residual connections, since this strategydoes not change the number of input and output channels of the convolutionblock. We propose a normalized distortion threshold to evaluate the sensitivityof each involved convolutional layer of the base model to guide STD-NET tocompress target network in an efficient and unsupervised approach, and obtaintwo network structures of different shapes with low computation cost andsimilar performance compared with the original one. Extensive experiments haveconfirmed that, on one hand, our model can achieve comparable or even betterdetection performance in various steganalytic scenarios due to the greatadaptivity of the obtained network architecture. On the other hand, theexperimental results also demonstrate that our proposed strategy is moreefficient and can remove more redundancy compared with previous steganalyticnetwork compression methods.

STD-NET: Search of Image Steganalytic Deep-learning Architecture via Hierarchical Tensor Decomposition