Tsunehiko Tanaka

Research Student

Tsunehiko Tanaka received his Bachelor of Engineering degree in 2021 from Waseda University and his Master of Engineering degree from the same university in 2023. He is currently pursuing his Ph.D. in machine learning at Waseda University. His primary research interests focus on the application of deep learning to games, including deep reinforcement learning and computer vision.

  • Lab: Simo-Serra Lab
  • Major at Univ: Computer Science
  • Email: tsunehiko@fuji.waseda.jp
  • City: Tokyo, Japan

Featured Project

Grammar-based Game Description Generation using Large Language Models

Tsunehiko Tanaka, Edgar Simo-Serra

This paper presents a novel framework that leverages Large Language Models (LLMs) to generate grammatically accurate game descriptions from natural language. [Software]

Transactions on Games, 2024

DiffG-RL: Leveraging Difference between State and Common Sense

Tsunehiko Tanaka, Daiki Kimura, Michiaki Tatsubori

A novel agent for text-based games, DiffG-RL, which properly organizes and handles common sense and environmental states. [Software]

Findings of EMNLP 2022

Commonsense Knowledge from Scene Graphs for Textual Environments

Tsunehiko Tanaka, Daiki Kimura, Michiaki Tatsubori

Application of common sense obtained from images to text-based games

AAAI 2022 Workshop (AAAI-RLG)

LoL-V2T: Large-Scale Esports Video Description Dataset

Tsunehiko Tanaka, Edgar Simo-Serra

A video captioning dataset for League of Legends [Project Page]

CVPR 2021 Workshop (CVSports)

Experience

Summary

Tsunehiko Tanaka

In university, I am studying computer science broadly, with a focus on machine learning. At the same time, I am actively participating in internships to learn industrial applications of the technologies I have learned.

  • tsunehiko@fuji.waseda.jp

Education

Doctor of Computer Science

2023.4 -

Waseda University, Tokyo, Japan

Engage in research in the area of reinforcement learning and game design under Prof. Simo-Serra.

Master of Computer Science

2021.4 - 2023.3

Waseda University, Tokyo, Japan

Engaged in research in the area of computer vision under Prof. Simo-Serra. Learned a variety of technologies related to machine learning, such as natural language processing and computer graphics.

Bachelor of Computer Science

2017.4 - 2021.3

Waseda University, Tokyo, Japan

Learned the basics of computer science such as algorithms, databases, etc.

Internship Experience

AI Research Intern

2023.7 -

CyberAgent AI Lab, Tokyo, Japan

  • Led a research project on offline reinforcement learning for AI-based game evaluation systems, managing the complete cycle from literature review to academic paper writing while building experimental environments with Docker on GCP.

Data Science Intern

2022.9 - 2023.3

SAP, Tokyo, Japan

  • Developed 3D point cloud processing technology to measure accumulated resource volumes at mining sites, creating a prototype for business development purposes.

AI Research Intern

2021.8 - 2022.5

IBM Research, Tokyo, Japan

  • Conducted research on reinforcement learning techniques for embodied learning using text-based games.

Machine Learning Engineer Intern

2021.2 - 2021.7

IDAJ Co., LTD., Yokohama, Japan

  • Surveyed and implemented video prediction models for anomaly detection in driving

Web Engineer Intern

2019.8 - 2021.1

Celsys.Inc., Tokyo, Japan

  • Improved community website for digital drawing software, CLIP STUDIO PAINT
  • Implemented bandit algorithms as benchmarking tools for recommendation engines

Publication

Preprint

  • Tsunehiko Tanaka, Kenshi Abe, Kaito Ariu, Tetsuro Morimura, Edgar Simo-Serra. 2024. Return-Aligned Decision Transformer. arXiv preprint, 2024. [arXiv]

Peer-reviewed Journal Papers

  • Tsunehiko Tanaka, Edgar Simo-Serra. 2024. Grammar-based Game Description Generation using Large Language Models. Transactions on Games. [DOI] [arXiv] [Software]

Peer-reviewed Conference Papers

  • Tsunehiko Tanaka, Daiki Kimura, Michiaki Tatsubori. 2022. DiffG-RL: Leveraging Difference between State and Common Sense. EMNLP 2022 Findings. [DOI] [Software]
  • Tsunehiko Tanaka, Daiki Kimura, Michiaki Tatsubori. 2022. Commonsense Knowledge from Scene Graphs for Textual Environments. AAAI-22 Workshop on Reinforcement Learning in Games. [Paper] [Software]
  • Daiki Kimura, Tsunehiko Tanaka, Michiaki Tatsubori, Asim Munawar. 2022. Challenges in Explainability and Knowledge Extraction. Wordplay: When Language Meets Games @ NAACL 2022. [Paper]
  • Tsunehiko Tanaka, Edgar Simo-Serra. 2021. LoL-V2T: Large-Scale Esports Video Description Dataset. 7th International Workshop on Computer Vision in Sports (CVsports) at CVPR2021. [DOI] [Project Page]

Domestic Conference Papers

  • Tsunehiko Tanaka, Edgar Simo-Serra. 2024. Object-Oriented Embeddings for Offline RL using Unsupervised Object Detection. 第 27 回画像の認識・理解シンポジウム(MIRU 2024).
  • Tsunehiko Tanaka, Edgar Simo-Serra. 2023. Leveraging Object Detectors for Online Action Detection. 第 26 回画像の認識・理解シンポジウム(MIRU 2023).
  • Tsunehiko Tanaka, Daiki Kimura, Michiaki Tatsubori. 2022. Reinforcement Learning with Common Sense from Visual Dataset. 第 25 回画像の認識・理解シンポジウム(MIRU 2022). (Short Oral, top 47%).
  • 田中恒彦, 木村大毅, 立堀道昭. 2022. シーングラフから抽出した常識のテキスト環境への応用. 人工知能学会全国大会(第 36 回). [DOI]
  • 木村 大毅, Chaudhury Subhajit, Swaminathan Sarathkrishna, 田中 恒彦, Agravante Don Joven, 立堀 道昭, Munawar Asim, Gray Alexander. 2022. ニューロシンボリック AI による強化学習. 人工知能学会全国大会(第 36 回). [DOI]
  • Tsunehiko Tanaka, Edgar Simo-Serra. 2021. LoL-V2T:Large-Scale Esports Video Description Dataset. 第 24 回画像の認識・理解シンポジウム(MIRU 2021).

Patents

  • Tsunehiko Tanaka, Daiki Kimura, Michiaki Tatsubori. Extracting enriched target-oriented common sense from grounded graphs to support next step decision making. U.S. Patent Application 17/812,757, filed 2024.

Research Grants and Projects

  • 科学技術振興機構 (JST) 戦略的創造研究推進事業 (Strategic Basic Research Programs) ACT-X, 人間とAIが理解できるインタラクション設計文法の構築 (Human and AI Understandable Interactive Design Grammars), PI. 2023.10-2026.3. [Page] [Abstract]
  • 日本学術振興会 (JSPS) 特別研究員 (Research Fellowship for Young Scientists, DC2), 人間と協力してデジタルコンテンツのインタラクションを設計するAIの構築 (Development of AI Systems that Collaborate with Humans in Designing Digital Content Interactions), PI. 2024.4-2026.3. [Page]
  • 科学技術振興機構 (JST) 次世代研究者挑戦的研究プログラム (Support for Pioneering Research Initiated by the Next Generation, SPRING), 人間と協力してデジタルコンテンツのインタラクションを設計するAIの構築 (Development of AI Systems that Collaborate with Humans in Designing Digital Content Interactions), PI. 2023.4-2024.3. [Page]

Awards

  • 大川功情報通信学術奨学金. 2024.
  • MIRU2023 若手プログラム. 発表“倫理審査・人を対象とした実験「AI研究者が配慮すべき倫理とは」” 優秀賞受賞. 2023. [Page]
  • 日本学生支援機構 第一種奨学金 全額返還免除. 2023.
  • 大川功情報通信学術奨学金. 2022.
  • MIRU2022 若手プログラム. 発表“研究費の取り方に関するサーベイ” 優秀賞受賞. 2022. [Page]
  • 日揮・実吉奨学会 奨学金. 2022.
  • JEES・ソフトバンク AI 人材育成奨学金. 2021.

Skills

  • Python, Java, JavaScript, PHP, Bash
  • Docker, Linux, PyTorch, Kubernates, Laravel