| Presentation | 2021-10-09 Explaining Adversarial Examples by the Embedding Structure of Data Manifold Hajime Tasaki, Yuji Kaneko, Jinhui Chao, |
|---|---|
| PDF Download Page |  PDF download Page Link |
| Abstract(in Japanese) | (See Japanese page) |
| Abstract(in English) | It is widely known that adversarial examples cause misclassification in classifiers using deep learning. Inspite of numerous previous research, the phenomenon has not been well understood and clearly explained, hence there is no effective countermeasures available now. In this research, instead of exploring on neural networks and their training algorithms, we focus on training data of the classifier, as a submanifold embedded in Euclidean space. We show that the occurrence of adversarial examples is due to the embedded structure of the data manifold in the data space. Particularly they lie in the complementary spaces of the tangent spaces of the data manifold. The theory is then verified by generation of adversarial examples and analysis of their distribution. |
| Keyword(in Japanese) | (See Japanese page) |
| Keyword(in English) | Adversarial Example / Deep Learning / Data Manifold |
| Paper # | PRMU2021-19 |
| Date of Issue | 2021-10-01 (PRMU) |
| Conference Information | |
| Committee | PRMU |
|---|---|
| Conference Date | 2021/10/8(2days) |
| Place (in Japanese) | (See Japanese page) |
| Place (in English) | Online |
| Topics (in Japanese) | (See Japanese page) |
| Topics (in English) | Processes and technologies to make research more efficient |
| Chair | Seiichi Uchida(Kyushu Univ.) |
| Vice Chair | Masakazu Iwamura(Osaka Pref. Univ.) / Mitsuru Anpai(Denso IT Lab.) |
| Secretary | Masakazu Iwamura(NTT) / Mitsuru Anpai(Tottori Univ.) |
| Assistant | Kouta Yamaguchi(CyberAgent) / Yusuke Matsui(Univ. of Tokyo) |
| Paper Information | |
| Registration To | Technical Committee on Pattern Recognition and Media Understanding |
|---|---|
| Language | JPN |
| Title (in Japanese) | (See Japanese page) |
| Sub Title (in Japanese) | (See Japanese page) |
| Title (in English) | Explaining Adversarial Examples by the Embedding Structure of Data Manifold |
| Sub Title (in English) | |
| Keyword(1) | Adversarial Example |
| Keyword(2) | Deep Learning |
| Keyword(3) | Data Manifold |
| 1st Author's Name | Hajime Tasaki |
| 1st Author's Affiliation | Chuo University(Chuo Univ.) |
| 2nd Author's Name | Yuji Kaneko |
| 2nd Author's Affiliation | Chuo University(Chuo Univ.) |
| 3rd Author's Name | Jinhui Chao |
| 3rd Author's Affiliation | Chuo University(Chuo Univ.) |
| Date | 2021-10-09 |
| Paper # | PRMU2021-19 |
| Volume (vol) | vol.121 |
| Number (no) | PRMU-192 |
| Page | pp.pp.17-21(PRMU), |
| #Pages | 5 |
| Date of Issue | 2021-10-01 (PRMU) |