Takeshi Koshizuka (越塚毅)

Takeshi Koshizuka

I am currently a Special Postdoctoral Researcher (SPDR) at the RIKEN Center for Advanced Intelligence Project (RIKEN AIP). I obtained my Ph.D. in Information Science and Technology from the University of Tokyo under the supervision of Prof. Issei Sato.

Google Scholar / Github / Twitter / CV

>> Japanese site

Publications

International Conferences (Peer-Reviewed)

[NeurIPS 2026] A Trust Region Approach for Learning Schrödinger Bridges
Takeshi Koshizuka, Denis Blessing, Max Zimmer, Sebastian Pokutta, Lorenz Richter
Advances in Neural Information Processing Systems 39 (NeurIPS 2026), Poster | [Paper (coming soon)]
[NeurIPS 2025] Understanding Generalization in Physics Informed Models through Affine Variety Dimensions
Takeshi Koshizuka, Issei Sato
Advances in Neural Information Processing Systems 38 (NeurIPS 2025), Poster | [arXiv]
[NeurIPS 2024] Understanding the Expressivity and Trainability of Fourier Neural Operator: A Mean-Field Perspective
Takeshi Koshizuka, Masahiro Fujisawa, Yusuke Tanaka, Issei Sato
Advances in Neural Information Processing Systems 37 (NeurIPS 2024), Poster | [arXiv]
[ICLR 2023] Neural Lagrangian Schrödinger Bridge: Diffusion Modeling for Population Dynamics
Takeshi Koshizuka, Issei Sato
International Conference on Learning Representations (ICLR 2023), Spotlight (notable-top-25%) | [Paper] [Video] [Code]
[Interspeech 2021] Fine-tuning Pre-trained Voice Conversion Model for Adding New Target Speakers with Limited Data
Takeshi Koshizuka, Hidefumi Ohmura, Kouichi Katsurada
Proceedings of Interspeech 2021, pp. 1339–1343 | [Paper] [Video]
[SuB 2020] A Neurolinguistic Investigation into Semantic Differences of Evidentiality and Modality
Yurie Hara, Naho Orita, Deng Ying, Takeshi Koshizuka, Sakai Hiromu
Proceedings of Sinn und Bedeutung 24, Vol. 1, pp. 273–290, 2020 | [Paper]

Journal Papers

[IPSJ Journal 2021] Graph-based Regional NMF for Distributed Computing
Takeshi Koshizuka, Koh Takeuchi, Tatsushi Matsubayashi, Hiroshi Sawada
IPSJ Journal, Vol. 62, No. 1, pp. 387–396, 2021 | [Paper]

Preprint

[arXiv 2026] Horizon-Aware Early Event Prediction for Tokamak Disruption Alarms
Takeshi Koshizuka, Takaharu Yaguchi
arXiv:2609.24443, 2026 | [arXiv]

Others (invited talks / domestic conferences)

[SCML 2026] Understanding Generalization in Physics Informed Models through Affine Variety Dimensions
Takeshi Koshizuka, Issei Sato
International Conference on Scientific Computing and Machine Learning (SCML 2026), University of Bath, UK, Sep 2026, Oral | [Conference]
[MIRU 2025] CARATTE: Camera and Radio-Based Attention Estimation
Takeshi Koshizuka, Ryo Eguchi, Matthew Ishige, Ryo Yonetani
Meeting on Image Recognition and Understanding (MIRU 2025), Demo presentation
[Invited Talk 2025] From Optimal Transport to Schrödinger Bridges: A Variational Perspective on Population Dynamics
Takeshi Koshizuka
REMODEL-DSC Workshop on Machine Learning and Numerical Analysis, Apr 2025, Invited talk
[IBIS 2023] Why Deep Fourier Neural Operator Performs Worse: Revisiting the Edge of Chaos
Takeshi Koshizuka, Masahiro Fujisawa, Yusuke Tanaka, Issei Sato
Information-Based Induction Sciences and Machine Learning Workshop (IBIS 2023), Nov 2023 | Best Student Presentation Award
[Invited Talk 2023] Diffusion Generative Model Learning Based on Schrödinger Bridge Problem
Takeshi Koshizuka
A New Era in Mathematical Science: The Synergy of Numerical Analysis and Machine Learning, Nov 2023, Invited talk
[IBIS 2022] Neural Lagrangian Schrödinger Bridge for Population Dynamics
Takeshi Koshizuka, Issei Sato
Information-Based Induction Sciences and Machine Learning Workshop (IBIS 2022), Nov 2022 | Outstanding Student Presentation Award
[IEICE 2021] A Vocoder-free Any-to-Many Voice Conversion using Pre-trained vq-wav2vec
Takeshi Koshizuka, Hidefumi Ohmura, Kouichi Katsurada
IEICE Technical Report, SP2021-03, Vol. 120, No. 399, pp. 176–181, Mar 2021 | [Paper]
[IEICE 2019] Evidentiality, Modality and Causality – Corpus and Neurolinguistic Studies
Yurie Hara, Naho Orita, Deng Ying, Takeshi Koshizuka, Sakai Hiromu
IEICE Technical Report, Vol. 119, No. 151, p. 15, 2019
[DICOMO 2019] Graph-based Regional NMF for Distributed Computing
Takeshi Koshizuka, Koh Takeuchi, Tatsushi Matsubayashi, Hiroshi Sawada
DICOMO 2019, pp. 743–751, 2019 | Best Research Paper Award

Education

D.Sc. in Comupter Science, The University of Tokyo, Apr. 2023 - Mar. 2026
M.Sc. in Computer Science, The University of Tokyo, Apr. 2021 - Mar. 2023

B.Sc. in Information Science, Tokyo University of Science, Apr. 2017 - Mar. 2021

Resarch & Work Experiences

JSPS Research Fellowships for Young Scientists DC1, Apr. 2023 - Mar. 2026
Visiting Researcher, ZIB, Germany, July 2025 - Dec 2025
Part-time Researcher, AI Lab @ CyberAgent, Inc. Japan, May. 2023 - May 2025

Data Scientist Internship, CyberAgent, Inc. Japan, Mar. 2023
Part-time Machine Learning Engineer, Future Architect, Inc. Japan, Oct. 2020 - Sep. 2021
Research Internship, NTT Communication Science Laboratories, Aug. 2018 - Sep 2018
Research Internship, NTT Communication Science Laboratories, Feb.2018 - Mar. 2018

Page updated

Google Sites

Report abuse