News

An article about my adoption of JSPS DC2 published on the website of the Department of Applied Informatics at Hosei University.

See also

• 北田俊輔さん (D2), 日本学術振興会特別研究員に.

An article about my adoption of JSPS DC2 published on the website of Hosei Univerisyt Graduate School.

See also

• 理工学研究科の北田俊輔氏が日本学術振興会特別研究員 (DC2) に内定.

• 北田 俊輔, 彌冨 仁. “半教師あり文書分類のための仮想敵対的学習による注意機構の頑健性および解釈性の向上”.
• 大畑 和也, 長澤 駿太, 北田 俊輔, 彌冨 仁. “キャプション生成時低品質データ事前検知の試み”.

Got Hosei University Research Grant for Doctoral Course Adopters in 2020 (480,000 JPY).

The following paper has been accepted to the AACL-IJCNLP 2020 Student Research Workshop (SRW).

• Takumi Aoki, Shunsuke Kitada, and Hitoshi Iyatomi. “Text Classification through Glyph-aware Disentangled Character Embedding and Semantic Sub-character Augmentation”

• 北田 俊輔, 彌冨 仁, 関 喜史. “生存時間分析を用いた広告クリエイティブの停止予測”.

Invited talk on Organized Session in The 34th Annual Conference of the Japanese Society for Artificial Intelligence.

See also

• The Present and Future of Machine Learning for Ad Creatives

The following paper has been accepted to the ACL2020 Student Research Workshop (SRW).

• Mahmoud Daif, Shunsuke Kitada, Hitoshi Iyatomi. “AraDIC: Arabic Document Classification using Image-Based Character Embeddings and Class-Balanced Loss”

• 青木 匠, 北田 俊輔, 彌冨 仁. “文字単位の解釈可能な潜在表現の data augmentation”.
• Mahmoud Daif, Shunsuke Kitada, Hitoshi Iyatomi. “Image-based Character Embedding for Arabic Document Classification”.
• 北田 俊輔, 彌冨 仁. “解釈可能な敵対的摂動を用いた頑健な注意機構の学習”.
• 長澤 駿太, 北田 俊輔, 彌冨 仁. “Script-aware embedding を用いた文字表現の獲得”.