| Presentation | 2023-03-16 Model compression by pruning of CNN based on perceptual hashes Shota Mishina, Tetsuya Morizumi, Hirotsugu Kinoshita, |
|---|---|
| PDF Download Page |  PDF download Page Link |
| Abstract(in Japanese) | (See Japanese page) |
| Abstract(in English) | Message digests that identify images are indispensable for secure and convenient copyright management of digital content. Considering the processing and editing of digital content during the distribution process, a conventional cryptographic hash function is not sufficient. Conventional perceptual hashing methods are not sufficiently resistant to processing and editing, but perceptual hashing based on convolutional neural network (CNN) weight coefficients is sufficiently resistant. However, the learned models must be shared between the generator and the verifier, and reducing the size of the shared data is a problem. In this study, we focus on pruning as a method of model compression for CNN models to reduce the shared data size, and evaluate the relationship between the accuracy of perceptual hash identification and the compression ratio. |
| Keyword(in Japanese) | (See Japanese page) |
| Keyword(in English) | Perceptual Hashing / Machine Learning / Pruning / Model Compression / Digital Copyright Management |
| Paper # | SITE2022-59,IA2022-82 |
| Date of Issue | 2023-03-08 (SITE, IA) |
| Conference Information | |
| Committee | IA / SITE / IPSJ-IOT |
|---|---|
| Conference Date | 2023/3/15(3days) |
| Place (in Japanese) | (See Japanese page) |
| Place (in English) | Maebashi Institute of Technology |
| Topics (in Japanese) | (See Japanese page) |
| Topics (in English) | Internet and Information Ethics Education, etc. |
| Chair | Tomoki Yoshihisa(Osaka Univ.) / Takushi Otani(Kibi International Univ.) |
| Vice Chair | Yusuke Sakumoto(Kwansei Gakuin Univ.) / Yuichiro Hei(KDDI Research) / Hiroshi Yamamoto(Ritsumeikan Univ.) / Soichiro Morishita(Cyber Agent) / Takeo Tatsumi(Open Univ. of Japan) |
| Secretary | Yusuke Sakumoto(Osaka Univ.) / Yuichiro Hei(Kogakuin Univ.) / Hiroshi Yamamoto(Kyushu Inst. of Tech.) / Soichiro Morishita(NRI-Secure) / Takeo Tatsumi(Hokuriku Univ.) |
| Assistant | Daisuke Kotani(Kyoto Univ.) / Ryo Nakamura(Fukuoka Univ.) / Ryo Nakamura(Univ. of Tokyo) / Yusuke Tachibana(Fukuoka Inst. of Tech.) |
| Paper Information | |
| Registration To | Technical Committee on Internet Architecture / Technical Committee on Social Implications of Technology and Information Ethics / Special Interest Group on Internet and Operation Technology |
|---|---|
| Language | JPN |
| Title (in Japanese) | (See Japanese page) |
| Sub Title (in Japanese) | (See Japanese page) |
| Title (in English) | Model compression by pruning of CNN based on perceptual hashes |
| Sub Title (in English) | |
| Keyword(1) | Perceptual Hashing |
| Keyword(2) | Machine Learning |
| Keyword(3) | Pruning |
| Keyword(4) | Model Compression |
| Keyword(5) | Digital Copyright Management |
| 1st Author's Name | Shota Mishina |
| 1st Author's Affiliation | Graduate School of Kanagawa University(Kanagawa Univ.) |
| 2nd Author's Name | Tetsuya Morizumi |
| 2nd Author's Affiliation | Kanagawa University(Kanagawa Univ.) |
| 3rd Author's Name | Hirotsugu Kinoshita |
| 3rd Author's Affiliation | Kanagawa University(Kanagawa Univ.) |
| Date | 2023-03-16 |
| Paper # | SITE2022-59,IA2022-82 |
| Volume (vol) | vol.122 |
| Number (no) | SITE-433,IA-434 |
| Page | pp.pp.28-34(SITE), pp.28-34(IA), |
| #Pages | 7 |
| Date of Issue | 2023-03-08 (SITE, IA) |