Shunsuke Kitada, Ph.D.

Shunsuke Kitada, Ph.D.

Research Scientist working on Vision & Language with Deep Learning

Research Scientist @ Image and Video AI Dept., LY Corp.

Kioi Tower 1-3 Kioicho, Chiyoda-ku, Tokyo, 102-8282, Japan

Project Researcher @ Hosei University

S603, 3-7-2, Kajino-cho, Koganei-shi, Tokyo 184-8584, Japan

Intelligent information processing laboratory

Biography

Shunsuke Kitada （北田 俊輔 in Japanese) is a Research Scientist at Image and Video AI Dept., LY Corp. His research interest is now on computational advertising with a focus on automatic geneartion/evaluation and assistive technology for multi-modal ad creatives.

Previously, he got his Ph.D. in 2023 at major in applied informatics, graduate school of science and engineering, Hosei University under the supervision of Prof. Hitoshi Iyatomi. His project was to improve prediction performance and model interpretability through attention mechanisms from basic and applied research perspectives.

His personality traits can be summarized as follows:

  • ❤️ Love research and development. He is enjoying research life, and he is currently focusing on three research topics simultaneously, e.g., natural language processing, medical image-based computer vision, computational advertising.
  • 📝 Every day read and implement the SoTA models. He has released many re-implementations of models using mainly Chainer and PyTorch. Therefore, based on state-of-the-art cases, he can advise on deep learning-based product design.
  • 😄 High technical communicativity. Summarize what he made and what he studied, and spread such information. This shows that he can input and output regularly.

The resume is available in PDF .

Background image borrowed from OpenAI ChatGPT illustration by Ruby Chen.

Interests
  • 🤖 Natural Language Processing
  • 💻 Computer Vision
  • 🏥 Medical Image Processing
  • 📃 Computational Advertising
Education

  • PhD in Engineering, 2023

    Graduate School of Science and Engineering, Hosei University

  • MSc in Engineering, 2020

    Graduate School of Science and Engineering, Hosei University

  • BSc in Engineering, 2018

    Hosei University

📢 Recent News

All news »

  • [2024.01] Published a special issue of doctoral dissertation in Journal of JSAI
  • [2023.03] Successfully Defended my Ph.D. Dissertation
  • [2023.01] Got Hosei University Research Grant 2022
  • [2022.11] Accepted our journal paper to IEEE Access
  • [2022.10] Accepted our journal paper to Springer Applied Intelligence journal
    • See all

    📝 Journal Article

    All Journal Articles »

    人工知能学会誌 博士論文集 2022.10～2023.9授与
    人工知能学会誌の博士論文特集号に寄稿しました。
    PDF DOI
    人工知能学会誌 博士論文集 2022.10～2023.9授与
    DM$^2$S$^2$: Deep Multi-Modal Sequence Sets with Hierarchical Modality Attention
    IEEE Access (Impact Factor: 3.476 in 2021; 1st place in Engineering & Computer Science (general) at Google Scholar Metrics)
    PDF Dataset DOI Preprint
    DM$^2$S$^2$: Deep Multi-Modal Sequence Sets with Hierarchical Modality Attention
    Making Attention Mechanisms More Robust and Interpretable with Virtual Adversarial Training
    Springer Applied Intelligence (Impact Factor: 5.019 in 2021)
    PDF DOI Preprint
    Making Attention Mechanisms More Robust and Interpretable with Virtual Adversarial Training
    Ad Creative Discontinuation Prediction with Multi-Modal Multi-Task Neural Survival Networks
    MDPI Appl. Sci. (Impact Factor: 2.838 in 2021; Top 10 ranking in Engineering & Computer Science (general) at Google Scholar Metrics)
    PDF DOI Preprint
    Ad Creative Discontinuation Prediction with Multi-Modal Multi-Task Neural Survival Networks
    Attention Meets Perturbations: Robust and Interpretable Attention with Adversarial Training
    IEEE Access (Impact Factor: 3.367 in 2020; 1st place in Engineering & Computer Science (general) at Google Scholar Metrics)
    PDF DOI Preprint Code
    Attention Meets Perturbations: Robust and Interpretable Attention with Adversarial Training

    📝 Conference Paper (Refereed)

    All Conference Papers »

    Majority or Minority: Data Imbalance Learning Method for Named Entity Recognition
    Practical ML for Low Resource Settings Workshop @ ICLR 2024; Rating: 9: Top 15% of accepted papers, strong accept; Oral presentation
    Preprint
    Majority or Minority: Data Imbalance Learning Method for Named Entity Recognition
    Feedback is Needed for Retakes: An Explainable Poor Image Notification Framework for the Visually Impaired
    Proc. IEEE 19th International Conference on Smart Communities: Improving Quality of Life Using ICT, IoT and AI (HONET).
    DOI Preprint
    Feedback is Needed for Retakes: An Explainable Poor Image Notification Framework for the Visually Impaired
    Expressions Causing Differences in Emotion Recognition in Social Networking Service Documents
    Proc. of the 31st ACM International Conference on Information & Knowledge Management (CIKM2022). (Acceptance rate = 29.04%)
    PDF Poster DOI Preprint Lightning talk
    Expressions Causing Differences in Emotion Recognition in Social Networking Service Documents
    Text Classification through Glyph-aware Disentangled Character Embedding and Semantic Sub-character Augmentation
    Proc. of the 1st Conference of the Asia-Pacific Chapter of the Association for Computational Linguistics and the 10th International Joint Conference on Natural Language Processing: Student Research Workshop (AACL-IJCNLP SRW 2020)
    PDF Preprint Code Video
    Text Classification through Glyph-aware Disentangled Character Embedding and Semantic Sub-character Augmentation
    AraDIC: Arabic Document Classification using Image-Based Character Embeddings and Class-Balanced Loss
    Proc. of the 58th Annual Meeting of the Association for Computational Linguistics: Student Research Workshop (ACL SRW 2020)
    PDF Video DOI Preprint Code
    AraDIC: Arabic Document Classification using Image-Based Character Embeddings and Class-Balanced Loss
    See all

    📝 Preprint

    All Preprints »

    Making Attention Mechanisms More Robust and Interpretable with Virtual Adversarial Training for Semi-Supervised Text Classification
    Preprint version of the Springer Applied Intelligence paper “Making Attention Mechanisms More Robust and Interpretable with Virtual Adversarial Training”
    DOI Preprint
    Making Attention Mechanisms More Robust and Interpretable with Virtual Adversarial Training for Semi-Supervised Text Classification
    Skin lesion classification with ensemble of squeeze-and-excitation networks and semi-supervised learning
    Manuscript for the International Skin Imaging Collaboration 2018.
    Preprint
    Skin lesion classification with ensemble of squeeze-and-excitation networks and semi-supervised learning

    🎓 Dissertation

    Improving Prediction Performance and Model Interpretability through Attention Mechanisms from Basic and Applied Research Perspectives
    Ph.D. dissertation
    Slides DOI Bulletin
    Improving Prediction Performance and Model Interpretability through Attention Mechanisms from Basic and Applied Research Perspectives

    🇯🇵 Domestic Conference / Presentation in Japan

    All domestic conference »

    根本 颯汰, 北田 俊輔, 彌冨 仁 (2023). Majority or Minority: 固有表現抽出におけるデータの不均衡性に着目した損失関数の提案. NLP 2023.

    PDF

    根本 颯汰, 北田 俊輔, 彌冨 仁 (2022). 固有表現認識タスクにおけるデータセットの偏りに着目した動的重み付け損失関数の提案. YANS 2022.

    北田 俊輔, 井上 直人, 大谷 まゆ, 彌冨 仁 (2022). Document AI タスクに向けた大規模事前学習済みモデルを活用した Layout-aware Prompting. YANS 2022.

    津嶋 祐介, 青木 匠, 北田 俊輔, 彌冨 仁 (2022). ViT-CLT: パッチ分割した文字画像から偏旁冠脚を考慮した文書分類. NLP 2022.

    PDF

    中川 翼, 北田 俊輔, 彌冨 仁 (2022). 書き手と読み手の怒りの感情認識の差の原因となる表現の獲得. NLP 2022.

    PDF

    中川 翼, 北田 俊輔, 彌冨 仁 (2021). 文章の書き手が持つ感情と読み手が受ける感情の差に関する初期調査および考察. YANS 2021.

    北田 俊輔, 岩崎 祐貴, 富樫 陸, 山口 光太, 彌冨 仁 (2021). 効果の高いランディングページ作成支援に向けた広告テキストからの効果予測および推薦モデルの構築. YANS 2021.

    北田 俊輔, 彌冨 仁, 関 喜史 (2021). 広告クリエイティブ評価のための深層確率埋め込みの学習. JSAI 2021.

    PDF JSAI

    大畑 和也, 長澤 駿太, 北田 俊輔, 彌冨 仁 (2021). キャプション生成時低品質データ事前検知の試み. NLP 2021.

    PDF

    北田 俊輔, 彌冨 仁 (2021). 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. NLP 2021.

    PDF

    See all

    🎙️ Recent & Upcoming Talks

    All recent and upcoming talks »

    BERT4CTR: An Efficient Framework to Combine Pre-trained Language Model with Non-textual Features for CTR Prediction
    In this talk, I presented the paper “BERT4CTR: An Efficient Framework to Combine Pre-trained Language Model with Non-textual Features for CTR Prediction”
    Jan 17, 2024 7:25 PM — 7:45 PM Online
    Slides
    BERT4CTR: An Efficient Framework to Combine Pre-trained Language Model with Non-textual Features for CTR Prediction
    Expressions Causing Differences in Emotion Recognition in Social Networking Service Documents (CIKM'22)
    In this talk, I presented the paper “Expressions Causing Differences in Emotion Recognition in Social Networking Service Documents” we presented at CIKM2022.
    Nov 12, 2022 2:45 PM — 3:00 PM Online
    Slides
    Expressions Causing Differences in Emotion Recognition in Social Networking Service Documents (CIKM'22)
    Introduction to AllenNLP to accelerate NLP research
    In this talk, I presented to introduce AllenNLP for accelerating NLP research in Japanese.
    Jul 22, 2022 7:30 PM — 8:30 PM Online
    Slides Video
    Introduction to AllenNLP to accelerate NLP research
    My experience about JSPS Research Fellowship
    In this seminar, I presented my experience as the only person who currently be adopted a JSPS DC2 in the campus.
    Apr 1, 2022 3:55 PM — 4:10 PM Online
    Slides
    My experience about JSPS Research Fellowship
    Recent Issues in Intelligent Robotics, Machine Learning, and Distributed System
    This webinar introduces advanced research projects on intelligent robotics, Machine Learning and Distributed System under the joint auspices of the engineering faculties of University of Carthage and Hosei University, IIST: Institute of Integrated Science and Technology, to promote international research and educational collaboration among their affiliated schools.
    Mar 17, 2021 6:45 PM — 7:55 PM Online
    Slides
    Recent Issues in Intelligent Robotics, Machine Learning, and Distributed System
    See all events

    🏆 Awards & Grants

  • [2023.01] Got Hosei University Research Grant 2022
  • [2022.07] Got JASSO Scholarship for Top 30% Excellent Graduate School Students
  • [2021.12] Got Hosei University Research Grant 2021
  • [2021.05] Got Grant-in-Aid for Japan Society for the Promotion of Science (JSPS) Fellows
  • [2021.04] Adopted as a JSPS DC2
    • See all

    📌 Featured Post

    All posts »

    Last updated on Jul 25, 2022 4 min read DC
    日本学術振興会特別研究員 (DC2) 申請書・学振
    日本学術振興会特別研究員 (DC2) 申請書・学振

    補欠繰り上げで 2021 年度学振特別研究員 DC2 に採用していただいた際の申請書です。

    📝 Posts & Articles

    About Awards and Interviews (Almost In 🇯🇵)

    画像生成 AI 入門：Python による拡散モデルの理論と実践
    https://coloso.jp/products/researchscientist-kitada-jp
    Last updated on Oct 9, 2023 1 min read Side job
    画像生成 AI 入門：Python による拡散モデルの理論と実践
    法政大学大学院入学案内 2023 - 学生・修了生の声
    https://saas.actibookone.com/?cNo=110942¶m=MV8zXzc=&pNo=1
    Last updated on May 16, 2022 2 min read Interview
    法政大学大学院入学案内 2023 - 学生・修了生の声
    人生が変わる研究インターンシップ 社会実装に触れて広がる研究キャリアとは
    https://www.cyberagent.co.jp/way/features/list/detail/id=26663
    Last updated on Oct 3, 2021 0 min read Interview
    人生が変わる研究インターンシップ 社会実装に触れて広がる研究キャリアとは
    北田俊輔さん (D2), 日本学術振興会特別研究員に
    https://ai.ws.hosei.ac.jp/wp/news/news20210408/
    Last updated on Jul 25, 2022 0 min read Interview
    北田俊輔さん (D2), 日本学術振興会特別研究員に
    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
    https://www.hosei.ac.jp/koganei/pickup/article-20210305121330/
    Last updated on Jul 25, 2022 0 min read Interview
    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
    See all posts

    👨‍💻 Selected Experiences

    All experiences can be found on my LinkedIn.

     
     
     
     
     
    Research Scientist
    LY Corporation
    October 2023 – Present Kioi Tower 1-3 Kioicho, Chiyoda-ku, Tokyo, 102-8282, Japan
    Transferred due to merge between LINE Corp. and Yahoo! Japan Corp.
     
     
     
     
     
    Research Scientist
    LINE Corporation
    April 2023 – September 2023 Yotsuya Tower 23rd FL., 1-6-1 Yotsuya, Shinjuku-ku, Tokyo, 160-0004, Japan
     
     
     
     
     
    Lecturer
    DAY1 COMPANY Inc.
    June 2023 – June 2026 Teheran-ro, Gangnam-gu, Seoul, Korea
    • Taught lectures on image-generating AI at Coloso, an online education service
    • The lectures can be found at the following URL: https://bit.ly/KitadaXColosoJP1
     
     
     
     
     
    Teaching Assistant for Deep Learning Hands-On Training Lab
    NVIDIA Japan
    January 2016 – April 2021

    Assisted in several workshops in Tokyo relating to deep learning and CUDA:

     
     
     
     
     
    Fundamental Information Technology Engineer
    Information-technology Promotion Agency, Japan
    June 2016 – Present
    Fundamental Information Technology Engineer Examination is a yardstick for measuring IT knowledge and skills as a team member by asking a range of questions about algorithm, network, database, information security, practical programming, etc.

    📂 Projects

    *
    All Official Implementation PyTorch Chainer Tool

    End-to-End Text Classification via Image-based Embedding using Character-level Networks

    Official Implementation of End-to-End Text Classification via Image-based Embedding using Character-level Networks

    Code Preprint IEEE Xplore

    AllenNLP Eraser

    Collection of AllenNLP DatasetReaders for ERASER

    Code
    Attention Meets Perturbation

    Attention Meets Perturbation

    Official Implementation of the paper Attention Meets Perturbations: Robust and Interpretable Attention with Adversarial Training

    Code Preprint IEEE Access

    Human Attention Map for Text Classification

    Reimplementation of the paper Human Attention Maps for Text Classification: Do Humans and Neural Networks Focus on the Same Words? (ACL2020) in PyTorch

    Code
    Paper Survey

    Paper Survey

    📚Survey of previous research and related works on machine learning (especially Deep Learning) in Japanese.

    Code
    Hosei University Soft Tennis Club Hp

    Hosei University Soft Tennis Club Hp

    🎾 Hosei University Soft Tennis Club’s home page repository.

    Follow Code

    chainer-FocalLoss

    Implementation of Focal loss in Chainer.

    Code

    chainer-RICAP

    Implementation of RICAP in Chainer.

    Code

    LSUV.pytorch

    Implementation of LSUV (Layer-sequential unit-variance) in PyTorch.

    Code Code

    libtorch-gin-api-server

    High-speed Deep learning API Server with Libtorch (C++) and Gin (Golang)

    Code

    nvhtop

    A tool for enriching the output of nvidia-smi forked from peci1/nvidia-htop.

    Code

    chainer-MeanTeachers

    Implementation of Mean teachers are better role models in Chainer.

    Code

    chainer-PyramidNet

    Implementation of PyramidNet in Chainer.

    Code

    chainer-InceptionResNetV2

    Implementation of InceptionResNetV2 in Chainer.

    Code

    chainer-IMSAT

    Implementation of IMSAT in Chainer.

    Code

    chainer-center-loss

    Implementation of Center Loss in Chainer.

    Code

    Conversion Prediction Using Multi-task Conditional Attention Networks to Support the Creation of Effective Ad Creatives

    Official Implementation of the paper Conversion Prediction Using Multi-task Conditional Attention Networks to Support the Creation of Effective Ad Creatives

    Code PDF

    📚 Popular Topics

    AllenNLP ANLP Awards & Grants Chainer Computational Advertising Domestic Conference Gunosy Hosei University International Conference International Publication Interview Invited talk Journal Natural Language Processing News Non-refereed PyTorch Refereed Referred YANS