Structural Image De-Identification for Privacy-Preserving Deep Learning
Due to the risk of data leakage while training deep learning models in a shared environment, we propose a new privacy-preserving deep learning (PPDL) method using a structural image de-identification approach for object classification. The proposed structural image de-identification approach is desi...
| Main Authors: | , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
IEEE
2020-01-01
|
| Series: | IEEE Access |
| Subjects: | |
| Online Access: | https://ieeexplore.ieee.org/document/9129656/ |
| _version_ | 1818665307067121664 |
|---|---|
| author | Dong-Hyun Ko Seok-Hwan Choi Jin-Myeong Shin Peng Liu Yoon-Ho Choi |
| author_facet | Dong-Hyun Ko Seok-Hwan Choi Jin-Myeong Shin Peng Liu Yoon-Ho Choi |
| author_sort | Dong-Hyun Ko |
| collection | DOAJ |
| description | Due to the risk of data leakage while training deep learning models in a shared environment, we propose a new privacy-preserving deep learning (PPDL) method using a structural image de-identification approach for object classification. The proposed structural image de-identification approach is designed based on the fact that the degree of structural distortion of an image object has the greatest impact on human's perceptual system. Thus, by modifying only the structural parts of the original one using order preserving encryption(OPE), the proposed structural image de-identification approach decreases only the recognition rate by human. From the experimental results using different standard datasets, we show that the object classification accuracy of the proposed structural image de-identification method is almost the same as the deep learning performance for non-encrypted images, without revealing the original image contents including sensitive information. Also, by handling the trade-off between object classification accuracy and privacy protection for the de-identified image, we experimentally find the optimal size of input image for the proposed structural image de-identification approach. |
| first_indexed | 2024-12-17T05:46:33Z |
| format | Article |
| id | doaj.art-85794be3897b42da88bdfc9504752f46 |
| institution | Directory Open Access Journal |
| issn | 2169-3536 |
| language | English |
| last_indexed | 2024-12-17T05:46:33Z |
| publishDate | 2020-01-01 |
| publisher | IEEE |
| record_format | Article |
| series | IEEE Access |
| spelling | doaj.art-85794be3897b42da88bdfc9504752f462022-12-21T22:01:17ZengIEEEIEEE Access2169-35362020-01-01811984811986210.1109/ACCESS.2020.30059119129656Structural Image De-Identification for Privacy-Preserving Deep LearningDong-Hyun Ko0https://orcid.org/0000-0002-5241-4702Seok-Hwan Choi1https://orcid.org/0000-0003-3590-6024Jin-Myeong Shin2Peng Liu3Yoon-Ho Choi4https://orcid.org/0000-0002-3556-5082School of Computer Science and Engineering, Pusan National University, Busan, South KoreaSchool of Computer Science and Engineering, Pusan National University, Busan, South KoreaSchool of Computer Science and Engineering, Pusan National University, Busan, South KoreaCollege of Information Sciences and Technology, Pennsylvania State University, State College, PA, USASchool of Computer Science and Engineering, Pusan National University, Busan, South KoreaDue to the risk of data leakage while training deep learning models in a shared environment, we propose a new privacy-preserving deep learning (PPDL) method using a structural image de-identification approach for object classification. The proposed structural image de-identification approach is designed based on the fact that the degree of structural distortion of an image object has the greatest impact on human's perceptual system. Thus, by modifying only the structural parts of the original one using order preserving encryption(OPE), the proposed structural image de-identification approach decreases only the recognition rate by human. From the experimental results using different standard datasets, we show that the object classification accuracy of the proposed structural image de-identification method is almost the same as the deep learning performance for non-encrypted images, without revealing the original image contents including sensitive information. Also, by handling the trade-off between object classification accuracy and privacy protection for the de-identified image, we experimentally find the optimal size of input image for the proposed structural image de-identification approach.https://ieeexplore.ieee.org/document/9129656/Data privacydeep learningimage encryptionstructural similarityvector graphics |
| spellingShingle | Dong-Hyun Ko Seok-Hwan Choi Jin-Myeong Shin Peng Liu Yoon-Ho Choi Structural Image De-Identification for Privacy-Preserving Deep Learning IEEE Access Data privacy deep learning image encryption structural similarity vector graphics |
| title | Structural Image De-Identification for Privacy-Preserving Deep Learning |
| title_full | Structural Image De-Identification for Privacy-Preserving Deep Learning |
| title_fullStr | Structural Image De-Identification for Privacy-Preserving Deep Learning |
| title_full_unstemmed | Structural Image De-Identification for Privacy-Preserving Deep Learning |
| title_short | Structural Image De-Identification for Privacy-Preserving Deep Learning |
| title_sort | structural image de identification for privacy preserving deep learning |
| topic | Data privacy deep learning image encryption structural similarity vector graphics |
| url | https://ieeexplore.ieee.org/document/9129656/ |
| work_keys_str_mv | AT donghyunko structuralimagedeidentificationforprivacypreservingdeeplearning AT seokhwanchoi structuralimagedeidentificationforprivacypreservingdeeplearning AT jinmyeongshin structuralimagedeidentificationforprivacypreservingdeeplearning AT pengliu structuralimagedeidentificationforprivacypreservingdeeplearning AT yoonhochoi structuralimagedeidentificationforprivacypreservingdeeplearning |