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

Deep image steganography is a data hiding technology that conceal data indigital images via deep neural networks. However, existing deep imagesteganography methods only consider the visual similarity of container imagesto host images, and neglect the statistical security (stealthiness) ofcontainer images. Besides, they usually hides data limited to image type andthus relax the constraint of lossless extraction. In this paper, we address theabove issues in a unified manner, and propose deep image steganography that canembed data with arbitrary types into images for secure data hiding and losslessdata revealing. First, we formulate the data hiding as an image colorizationproblem, in which the data is binarized and further mapped into the colorinformation for a gray-scale host image. Second, we design a conditionalinvertible neural network which uses gray-scale image as prior to guide thecolor generation and perform data hiding in a secure way. Finally, to achievelossless data revealing, we present a multi-stage training scheme to manage thedata loss due to rounding errors between hiding and revealing processes.Extensive experiments demonstrate that the proposed method can perform securedata hiding by generating realism color images and successfully resisting thedetection of steganalysis. Moreover, we can achieve 100% revealing accuracy indifferent scenarios, indicating the practical utility of our steganography inthe real-world.

Hiding Data in Colors: Secure and Lossless Deep Image Steganography via Conditional Invertible Neural Networks