Yuki Kadokawa.
Portrait of Yuki Kadokawa

Yuki Kadokawa

Sim-to-Real Reinforcement Learning for Robotics

I am a researcher in reinforcement learning (RL), deep learning, and robotics. My research focuses on Sim-to-Real transfer — how robots can learn in simulation and operate in the real world — using methods such as domain randomization, policy distillation, and sample-efficient reinforcement learning. I have worked on real-world robotic tasks including manipulation, legged locomotion, multi-robot systems, and field and industrial robotics.

Cross-appointment
  • Researcher, Fukushima Institute for Research, Education and Innovation (F-REI)
  • Project Assistant Professor, Nara Institute of Science and Technology (NAIST) — Robot Learning Lab
Google Scholar GitHub

News & Highlights

2026
NEWGave an invited talk at the SICE Symposium on Decentralized Autonomous Systems (自律分散システム・シンポジウム) on Sim-to-Real RL for real-world robot tasks.
2026
DAPPER was accepted to IEEE Robotics & Automation Magazine (RAM) and presented at ICRA 2026 — joint work with ETH Zürich.
2026
Our paper on Distilled Iterative Value Conversion for neurochip-driven edge robots was published in IEEE Access.
2025
Progressive-Resolution Policy Distillation (PRPD) was published in IEEE T-ASE and featured in the Nikkan Kogyo Shimbun.
2025
Learning Quiet Walking for a Small Home Robot was presented at ICRA 2025 — collaboration with ETH Zürich and Sony.
2024
Completed my Doctor of Engineering at NAIST and began as a Project Assistant Professor in the Robot Learning Lab.

Publications

27

International Journal

Figure 1 of Distilled Iterative Value Conversion Video thumbnail
IEEE Access2026
Distilled Iterative Value Conversion: Reducing FPNN-to-SNN Conversion Errors via Distillation in Reinforcement Learning for Neurochip-Driven Edge Robots

Yuki Kadokawa, Tomoya Yamanokuchi, Alonso Ramos Fernandez, Takanori Homma, and Takamitsu Matsubara

IEEE Access, 2026

Paper Video
Figure 1 of DAPPER Video thumbnail
IEEE RAM2026
DAPPER: Discriminability-Aware Policy-to-Policy Preference-Based Reinforcement Learning for Query-Efficient Robot Skill Acquisition

Yuki Kadokawa, Jonas Frey, Takahiro Miki, Takamitsu Matsubara, and Marco Hutter

IEEE Robotics & Automation Magazine (RAM), 2026

Paper Video
Figure 1 of Prolonging Tool Life
IEEE Access2026
Prolonging Tool Life: Learning Skillful Use of General-purpose Tools through Lifespan-guided Reinforcement Learning

Po-Yen Wu, Cheng-Yu Kuo, Yuki Kadokawa, and Takamitsu Matsubara

IEEE Access, 2026

Paper Project Page
Figure 1 of Progressive-Resolution Policy Distillation Video thumbnail
IEEE T-ASE2025
Progressive-Resolution Policy Distillation: Leveraging Coarse-Resolution Simulations for Time-Efficient Fine-Resolution Policy Learning

Yuki Kadokawa, Hirotaka Tahara, and Takamitsu Matsubara

IEEE Transactions on Automation Science and Engineering (T-ASE), 2025

Paper Project Page Video Newspaper
Figure 1 of Robust Iterative Value Conversion Video thumbnail
RAS2024
Robust Iterative Value Conversion: Deep Reinforcement Learning for Neurochip-driven Edge Robots

Yuki Kadokawa, Tomohito Kodera, Yoshihisa Tsurumine, Shinya Nishimura, and Takamitsu Matsubara

Robotics and Autonomous Systems (RAS), 2024

Paper Video
Figure 1 of Cyclic Policy Distillation Video thumbnail
RAS2023
Cyclic Policy Distillation: Sample-Efficient Sim-to-Real Reinforcement Learning with Domain Randomization

Yuki Kadokawa, Lingwei Zhu, Yoshihisa Tsurumine, and Takamitsu Matsubara

Robotics and Autonomous Systems (RAS), 2023

Paper Video
Figure 1 of Binarized P-Network Video thumbnail
IEEE RA-L2021
Binarized P-Network: Deep Reinforcement Learning of Robot Control from Raw Images on FPGAStudent Paper Award

Yuki Kadokawa, Yoshihisa Tsurumine, and Takamitsu Matsubara

IEEE Robotics and Automation Letters (RA-L), vol. 6, no. 4, pp. 8545–8552, 2021

Paper Video
Figure 1 of ViSA
Under Review
ViSA: Visited-State Augmentation for Generalized Goal-Space Contrastive Reinforcement Learning

Issa Nakamura, Tomoya Yamanokuchi, Yuki Kadokawa, Jia Qu, Shun Otsubo, Shotaro Miwa, and Takamitsu Matsubara

Under review

Paper
Figure 1 of Autonomous Obstacle Removal for Excavators Video thumbnail
Under Review
Autonomous Obstacle Removal for Excavators through Policy Learning with Particle Simulation

Yuki Kadokawa, Sandro M. Alcantara Tacora, Taro Abe, Daisuke Endo, Genki Yamauchi, Takeshi Hashimoto, and Takamitsu Matsubara

Under review

Paper Video
Figure 1 of Bridged SBI
Under Review
Bridged SBI: Correcting Biased Low-Fidelity Posteriors for Cost-Efficient High-Fidelity Inference

Gahee Kim, Yuki Kadokawa, Sandro M. Alcantara Tacora, Taro Abe, Daisuke Endo, Genki Yamauchi, Takeshi Hashimoto, and Takamitsu Matsubara

Under review

Paper

International Conference

ICRA2026
DAPPER: Discriminability-Aware Policy-to-Policy Preference-Based Reinforcement Learning for Query-Efficient Robot Skill Acquisition

Yuki Kadokawa, Jonas Frey, Takahiro Miki, Takamitsu Matsubara, and Marco Hutter

IEEE International Conference on Robotics and Automation (ICRA, RAM option), 2026

Paper Video
ICRA2026
Progressive-Resolution Policy Distillation: Leveraging Coarse-Resolution Simulations for Time-Efficient Fine-Resolution Policy Learning

Yuki Kadokawa, Hirotaka Tahara, and Takamitsu Matsubara

IEEE International Conference on Robotics and Automation (ICRA, T-ASE option), 2026

Paper Project Page Video Newspaper
Figure 1 of Robust Sim-to-Real Cloth Untangling Video thumbnail
CASE2026
Robust Sim-to-Real Cloth Untangling through Reduced-Resolution Observations via Adaptive Force-Difference Quantization

Yoshihisa Tsurumine, Yuki Kadokawa, Kohei Hayashi, Christian Diehm, and Takamitsu Matsubara

IEEE International Conference on Automation Science and Engineering (CASE), 2026

Paper Video
Figure 1 of Learning Quiet Walking for a Small Home Robot Video thumbnail
ICRA2025
Learning Quiet Walking for a Small Home Robot

Ryo Watanabe, Takahiro Miki, Fan Shi, Yuki Kadokawa, Filip Bjelonic, Kento Kawaharazuka, Andrei Cramariuc, and Marco Hutter

International Conference on Robotics and Automation (ICRA), 2025

Paper Project Page Video
AROB2025
Scalable Domain Randomized Reinforcement Learning for Sim-to-Real Policy Transfer in Complex Robot Tasks

Yuki Kadokawa, and Takamitsu Matsubara

International Symposium on Artificial Life and Robotics (AROB), 2025

Conference
Figure 1 of Learning Robotic Powder Weighing from Simulation Video thumbnail
IROS2023
Learning Robotic Powder Weighing from Simulation for Laboratory Automation

Yuki Kadokawa, Masashi Hamaya, and Kazutoshi Tanaka

IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS), 2023

Paper Project Page Video
ICRA2023
Binarized P-Network: Deep Reinforcement Learning of Robot Control from Raw Images on FPGA

Yuki Kadokawa, Yoshihisa Tsurumine, and Takamitsu Matsubara

IEEE International Conference on Robotics and Automation (ICRA, RA-L option), 2023

Paper Video
Figure 1 of DeReCo Video thumbnail
Under Review
DeReCo: Decoupling Representation and Coordination Learning for Object-Adaptive Decentralized Multi-Robot Cooperative Transport

Kazuki Shibata, Ryosuke Sota, Shandil Dhiresh Bosch, Yuki Kadokawa, Yoshihisa Tsurumine, and Takamitsu Matsubara

Under review

Paper Video

Domestic Conference

SICE SI2025
段階的解像度シミュレーションを用いた土砂の挙動を再現する粒子パラメータ自動推定

金加喜, 角川勇貴, Sandro Manuel Alcantara Tacora, 阿部太郎, 遠藤大輔, 山内元貴, 橋本毅, 松原崇充

計測自動制御学会システムインテグレーション部門講演会, 2025

Conference
RSJ2025
ドメインランダム化強化学習による油圧ショベルの土砂中の障害物除去タスク実現

角川勇貴, Sandro Manuel Alcantara Tacora, 松原崇充

日本ロボット学会学術講演会, 2025

Conference
RSJ2025
到達状態拡張による対照強化学習の汎化

中村維冴, 山之口智也, 角川勇貴, 曲佳, 大坪舜, 宮本健, 三輪祥太郎, 松原崇充

日本ロボット学会学術講演会, 2025

Conference
RSJ2025
柔軟な群ロボット協調輸送のための非対称アクター・クリティック型マルチエージェント強化学習

曽田涼介, 柴田一騎, 鶴峯義久, 角川勇貴, 松原崇充

日本ロボット学会学術講演会, 2025

Conference
RSJ2024
Deep Reinforcement Learning with FPNN-to-SNN Policy Distillation for Neurochip-driven RobotsBest Presentation Award Finalist

Alonso Ramos Fernandez, Yuki Kadokawa, Yoshihisa Tsurumine, and Takamitsu Matsubara

Annual Conference of the Robotics Society of Japan, 2024

Conference
ROBOMECH2024
Sim-to-Real 方策転移のためのダイナミクスランダム化対照強化学習

中村維冴, 山之口智也, 角川勇貴, 曲佳, 大坪舜, 三輪祥太郎, 松原崇充

ロボティクス・メカトロニクス講演会, 2024

Conference
SICE MSCS2023
ニューロチップ実装に適した量子化方策のエッジサーバー深層強化学習

小寺智仁, 角川勇貴, 鶴峯義久, 松原崇充

計測自動制御学会制御部門マルチシンポジウム, 2023

Conference
RSJ2020
FPGAを用いた実時間ロボット制御のための深層強化学習手法 Binary P-Network の提案

角川勇貴, 鶴峯義久, 松原崇充

日本ロボット学会学術講演会, 2020

Conference

Invited Talk

SICE DAS2026
実世界ロボットタスクにおけるSim-to-Real強化学習: パーティクルシミュレーションにおける計算量と精度のトレードオフ

角川勇貴, 松原崇充

計測自動制御学会 自律分散システム・シンポジウム, 2026

Conference

Honors

Awards

  1. Outstanding Student Award, Symposium of University Fellowship, NAIST, 2022 [Award Page]
  2. Student Paper Award, IEEE Kansai Section, 2022 [NAIST] [IEEE]
  3. Best Student of the Year, Toyama Prefectural University, 2019
  4. Grand Prize, Monozukuri in Toyama, Toyama Mechanical and Electrical Industries Association, 2018
  5. Best Presentation Award, Design Competition, Japan Society for Design Engineering, 2018
  6. 3rd Place (National Finals 3/10), Design Competition, Japan Society for Design Engineering, 2018
  7. Best Student Presentation Award, Research Presentation, Japan Society for Design Engineering, 2018
  8. Best Presentation Award, Design Competition, Japan Society for Design Engineering, 2017
  9. Best Student of the Year, Toyama Prefectural Takaoka Kogei High School, 2015

Scholarship & Grant

  1. Senju Monju Project (Research Grant), NAIST, 2024–2026
  2. Scholarship (Return Exemption, Full Amount), Japan Student Services Organization, 2023
  3. Research Fellowship for Young Scientists: DC2 (Funding & Salary), JSPS, 2023–2025
  4. Research Fellowship (Research Grant & Salary), Japan Science and Technology Agency, 2022–2023 [Interview]
  5. Scholarship (Return Exemption, Half Amount), Japan Student Services Organization, 2021
  6. Outstanding Student Scholarship (Tuition Exemption), NAIST, 2021

Experience & Education

Work History

  • Apr 2024 – Mar 2026
    Project Assistant Professor
    Robot Learning Lab, NAIST
  • Aug 2023 – Oct 2023
    Visiting Researcher
    Robotic Systems Lab, ETH Zürich
  • Apr 2022 – Mar 2023
    Research Internship
    OMRON SINIC X Corporation
    Project Page Video
  • Dec 2018 – Mar 2019
    Part-time, 3D-CAD Modeling
    Aluminum Factory Corporation

Education

  • 2024
    Doctor of Engineering
    Program of Information Science and Engineering, NAIST, Japan
  • 2021
    Master of Engineering
    Program of Information Science and Engineering, NAIST, Japan
  • 2019
    Bachelor of Engineering
    Department of Intelligent Robotics, Toyama Prefectural University, Japan

Research Field

Reinforcement Learning Deep Learning Robotics Sim-to-Real Domain Randomization Edge Robots FPGA Neurochip

Contact

E-mail
kadokawa.yuki [at] naist.ac.jp
Address
8916-5 Takayama-cho, Ikoma, Nara 630-0192, Japan
Affiliation
Nara Institute of Science and Technology, Robot Learning Laboratory
Google Scholar GitHub