Hiroshi Takahashi

Constraining Logits by Bounded Function for Adversarial Robustness pdf

Hiroshi Takahashi, Yasutoshi Ida, Sekitoshi Kanai, 2020, 2021 International Joint Conference on Neural Networks (IJCNN).

One-vs-the-Rest Loss to Focus on Important Samples in Adversarial Training

Yasutoshi Ida, Sekitoshi Kanai, Shin'ya Yamaguchi, 2022, ICML.

Switching One-Versus-the-Rest Loss to Increase the Margin of Logits for Adversarial Robustness

Yasutoshi Ida, Sekitoshi Kanai, Shin'ya Yamaguchi, 2022, ArXiv.

Smoothness Analysis of Adversarial Training

Yasutoshi Ida, Sekitoshi Kanai, Masanori Yamada, 2021, 2103.01400.

Adversarial Training Makes Weight Loss Landscape Sharper in Logistic Regression pdf

Tomoharu Iwata, Atsutoshi Kumagai, Sekitoshi Kanai, 2021, ArXiv.

S WITCHING O NE -V ERSUS - THE -R EST L OSS TO I NCREASE L OGIT M ARGINS FOR A DVERSARIAL R OBUSTNESS

Sekitoshi Kanai, Shin'ya Yamaguchi, Masanori Yamada, 2022 .

Relationship Between Nonsmoothness in Adversarial Training, Constraints of Attacks, and Flatness in the Input Space

Yasutoshi Ida, Sekitoshi Kanai, Masanori Yamada, 2023, IEEE Transactions on Neural Networks and Learning Systems.