HiDDeN: Hiding Data With Deep Networks

  title={HiDDeN: Hiding Data With Deep Networks},
  author={Jiren Zhu and Russell Kaplan and Justin Johnson and Li Fei-Fei},
Recent work has shown that deep neural networks are highly sensitive to tiny perturbations of input images, giving rise to adversarial examples. Though this property is usually considered a weakness of learned models, we explore whether it can be beneficial. We find that neural networks can learn to use invisible perturbations to encode a rich amount of useful information. In fact, one can exploit this capability for the task of data hiding. We jointly train encoder and decoder networks, where… CONTINUE READING
Related Discussions
This paper has been referenced on Twitter 33 times. VIEW TWEETS


Publications citing this paper.
Showing 1-2 of 2 extracted citations


Publications referenced by this paper.
Showing 1-10 of 31 references

Designing steganographic distortion using directional filters

2012 IEEE International Workshop on Information Forensics and Security (WIFS) • 2012
View 5 Excerpts
Highly Influenced

Similar Papers

Loading similar papers…