DSteganoM: Deep Steganography for Motion Capture Data

Qi Wen Gan1, Wei-Chuen Yau1, Y.S. Gan2, Iftekhar Salam1, Shihui Guo3, Chin-Chen Chang2, Yubing Wu1 Luchen Zhou1
1Xiamen University Malaysia, 2Feng Chia University, 3Xiamen University

One-sentence summary: Hiding Mocap Data from one Skeleton Animation to another Skeleton Animation using end-to-end Deep Neural Network.



⚠️

Please click on each of the following pictures to view the corresponding sample animation.

Sample One : Hiding one Human Skeleton Animation inside another Human Skeleton Animation

CMU - Our Method

Cover

Secret

Corrupted Setting

CMU - SkipGated

Cover

Secret

Corrupted Setting

Sample Two: Hiding one Hand Skeleton Animation inside another Hand Skeleton Animation

MTM - Our Method

Cover

Secret

Corrupted Setting

MTM - SkipGated

Cover

Secret

Corrupted Setting

BibTeX

@article{DSteganoM2022,
  author    = {Qi Wen, Gan and Wei-Chuen, Yau and Y.S., Gan and Iftekhar, Salam and Shihui, Guo and Chin-Chen, Chang and Yubing, Wu and Luchen, Zhou},
  title     = {DSteganoM: Deep Steganography for Motion Capture Data},
  year      = {2022},
}