Digital Hologram Watermarking Based on Multiple Deep Neural Networks Training Reconstruction and Attack
This paper proposes a method to embed and extract a watermark on a digital hologram using a deep neural network. The entire algorithm for watermarking digital holograms consists of three sub-networks. For the robustness of watermarking, an attack simulation is inserted inside the deep neural network...
| Main Authors: | , , , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
MDPI AG
2021-07-01
|
| Series: | Sensors |
| Subjects: | |
| Online Access: | https://www.mdpi.com/1424-8220/21/15/4977 |
| _version_ | 1797525208995725312 |
|---|---|
| author | Ji-Won Kang Jae-Eun Lee Jang-Hwan Choi Woosuk Kim Jin-Kyum Kim Dong-Wook Kim Young-Ho Seo |
| author_facet | Ji-Won Kang Jae-Eun Lee Jang-Hwan Choi Woosuk Kim Jin-Kyum Kim Dong-Wook Kim Young-Ho Seo |
| author_sort | Ji-Won Kang |
| collection | DOAJ |
| description | This paper proposes a method to embed and extract a watermark on a digital hologram using a deep neural network. The entire algorithm for watermarking digital holograms consists of three sub-networks. For the robustness of watermarking, an attack simulation is inserted inside the deep neural network. By including attack simulation and holographic reconstruction in the network, the deep neural network for watermarking can simultaneously train invisibility and robustness. We propose a network training method using hologram and reconstruction. After training the proposed network, we analyze the robustness of each attack and perform re-training according to this result to propose a method to improve the robustness. We quantitatively evaluate the results of robustness against various attacks and show the reliability of the proposed technique. |
| first_indexed | 2024-03-10T09:09:29Z |
| format | Article |
| id | doaj.art-c52cdc449dd74d4fbe8af3b168e95c61 |
| institution | Directory Open Access Journal |
| issn | 1424-8220 |
| language | English |
| last_indexed | 2024-03-10T09:09:29Z |
| publishDate | 2021-07-01 |
| publisher | MDPI AG |
| record_format | Article |
| series | Sensors |
| spelling | doaj.art-c52cdc449dd74d4fbe8af3b168e95c612023-11-22T06:08:37ZengMDPI AGSensors1424-82202021-07-012115497710.3390/s21154977Digital Hologram Watermarking Based on Multiple Deep Neural Networks Training Reconstruction and AttackJi-Won Kang0Jae-Eun Lee1Jang-Hwan Choi2Woosuk Kim3Jin-Kyum Kim4Dong-Wook Kim5Young-Ho Seo6Department of Electronic Materials Engeering, Kwangwoon University, Kwangwoon-ro 20, Nowon-gu, Seoul 01897, KoreaOLED Team Associate, Siliconworks, Baumoe-ro, Seocho-gu, Seoul 06763, KoreaDepartment of Electronic Materials Engeering, Kwangwoon University, Kwangwoon-ro 20, Nowon-gu, Seoul 01897, KoreaDepartment of Electronic Materials Engeering, Kwangwoon University, Kwangwoon-ro 20, Nowon-gu, Seoul 01897, KoreaDepartment of Electronic Materials Engeering, Kwangwoon University, Kwangwoon-ro 20, Nowon-gu, Seoul 01897, KoreaDepartment of Electronic Materials Engeering, Kwangwoon University, Kwangwoon-ro 20, Nowon-gu, Seoul 01897, KoreaDepartment of Electronic Materials Engeering, Kwangwoon University, Kwangwoon-ro 20, Nowon-gu, Seoul 01897, KoreaThis paper proposes a method to embed and extract a watermark on a digital hologram using a deep neural network. The entire algorithm for watermarking digital holograms consists of three sub-networks. For the robustness of watermarking, an attack simulation is inserted inside the deep neural network. By including attack simulation and holographic reconstruction in the network, the deep neural network for watermarking can simultaneously train invisibility and robustness. We propose a network training method using hologram and reconstruction. After training the proposed network, we analyze the robustness of each attack and perform re-training according to this result to propose a method to improve the robustness. We quantitatively evaluate the results of robustness against various attacks and show the reliability of the proposed technique.https://www.mdpi.com/1424-8220/21/15/4977digital hologramdigital watermarkdeep neural network (DNN)training datasetconvolution neural network (CNN) |
| spellingShingle | Ji-Won Kang Jae-Eun Lee Jang-Hwan Choi Woosuk Kim Jin-Kyum Kim Dong-Wook Kim Young-Ho Seo Digital Hologram Watermarking Based on Multiple Deep Neural Networks Training Reconstruction and Attack Sensors digital hologram digital watermark deep neural network (DNN) training dataset convolution neural network (CNN) |
| title | Digital Hologram Watermarking Based on Multiple Deep Neural Networks Training Reconstruction and Attack |
| title_full | Digital Hologram Watermarking Based on Multiple Deep Neural Networks Training Reconstruction and Attack |
| title_fullStr | Digital Hologram Watermarking Based on Multiple Deep Neural Networks Training Reconstruction and Attack |
| title_full_unstemmed | Digital Hologram Watermarking Based on Multiple Deep Neural Networks Training Reconstruction and Attack |
| title_short | Digital Hologram Watermarking Based on Multiple Deep Neural Networks Training Reconstruction and Attack |
| title_sort | digital hologram watermarking based on multiple deep neural networks training reconstruction and attack |
| topic | digital hologram digital watermark deep neural network (DNN) training dataset convolution neural network (CNN) |
| url | https://www.mdpi.com/1424-8220/21/15/4977 |
| work_keys_str_mv | AT jiwonkang digitalhologramwatermarkingbasedonmultipledeepneuralnetworkstrainingreconstructionandattack AT jaeeunlee digitalhologramwatermarkingbasedonmultipledeepneuralnetworkstrainingreconstructionandattack AT janghwanchoi digitalhologramwatermarkingbasedonmultipledeepneuralnetworkstrainingreconstructionandattack AT woosukkim digitalhologramwatermarkingbasedonmultipledeepneuralnetworkstrainingreconstructionandattack AT jinkyumkim digitalhologramwatermarkingbasedonmultipledeepneuralnetworkstrainingreconstructionandattack AT dongwookkim digitalhologramwatermarkingbasedonmultipledeepneuralnetworkstrainingreconstructionandattack AT younghoseo digitalhologramwatermarkingbasedonmultipledeepneuralnetworkstrainingreconstructionandattack |