Kuniyuki Takahashi,Ph.D
Preferred Networks, Engineering Manager & Researcher
Email: kuniyuki.takahashi[at]gmail.com
Kuniyuki Takahashi,Ph.D
Preferred Networks, Engineering Manager & Researcher
Email: kuniyuki.takahashi[at]gmail.com
Robotics researcher with 15+ years of experience, specializing in deep learning for robotics and real-world problem solving.
Currently leading research and development of Vision Language Models (VLMs), covering model design, training, evaluation, and application.
Lead R&D of VLMs: design, training, evaluation, and application
Develop deep learning methods for real-world robot manipulation
Award-winning research at top-tier conferences/journals (e.g., ICRA; Advanced Robotics)
Inventor on multiple US/JP patents (see the Patents section for details)
Deep Learning for Robotics
Multi-modal Learning and Sensor Fusion
Reinforcement Learning
Computer Vision
Vision Language Model (VLM)
Contact-rich Manipulation: manipulation involving physical contact; adaptive learning under uncertainty
Multimodal Learning: sensor fusion centered on vision and tactile sensing
Data-efficient Learning: learning with limited data ; synthetic data generation, etc.
Flexible Object / Flexible Joint Dynamics: manipulation of deformable objects; motion generation for robots with flexible joints
Robotics × Language: design, training, evaluation, and application of models integrating vision and language
Vision Language Model (VLM): Robust models for VQA, visual grounding, task analysis, and anomaly detection.
2024 FA Foundation Paper Award (Co-author)
2024 Advanced Robotics Best Paper Award (Co-author)
2019 FA Foundation Paper Award (First author)
2019 Advanced Robotics Best Paper Award (First author)
2019 RSJ2019 Young Investigation Excellence Award (First author)
2019 ICRA2019 Best Paper Award finalist (First author)
2018 ICRA2018 Best Paper Award on HRI (Equal contribution first author)
2016 SI2016 Award (Co-author)
2016 IPSJ2016 Student Encouragement Award (Co-author)
2016 IPSJ2016 Conference Encouragement Award (Co-author)
2011 SI2011 Award (Presenter)
Engineering Manager (20th January 2025 - Current)
Researcher (1st April 2017 - Current)
Secondment; Preferred Elements, Inc. (20th January 2025 - 30th June 2025)
Ristumeika University
Visiting Researcher (1st November 2021 - 31st March 2024)
Waseda University
Adjunct Researcher (1st December 2017 - Current)
Institute for Cognitive Systems (Prof. Gordon Cheng), Technical University of Munich, Munich, Germany
Visiting researcher (February 2017 - March 2017)
Visiting researcher (February 2015 - March 2016)
Internship (4th July 2016 - 29th July 2016)
Japan Society for the Promotion of Science
Research Fellow DC2 (April 2015 - March 2017)
Waseda University
Graduate Program for Embodiment Informatics (February 2014 - March 2017)
Research Assistant (RA) (September 2013 - April 2014)
Research assistant in the brain circulation Abroad Program (April 2013 - July 2013)
Teaching Assistant(TA) (April 2011 - March 2012)
April 2013 - March 2017
Doctor of Engineering, Department of Modern Mechanical Engineering, Graduate School of Creative Science and Engineering, Waseda University
Ph.D. thesis: “Acquisition and Extension of the Robot Body Model Using Motor Babbling through Deep Learning” [PDF]
Supervisor: Prof. Shigeki Sugano
Co-adviser: Prof. Tetsuya Ogata (Waseda University, Japan)
Co-adviser: Prof. Gordon Cheng (Technical University of Munich, Germany)
April 2011 - March 2013
Master of Engineering, Department of Modern Mechanical Engineering, Graduate School of Creative Science and Engineering, Waseda University
April 2007 - March 2011
Bachelor of Engineering, Department of Modern Mechanical Engineering, School of Creative Science and Engineering, Waseda University
February 2015 - March 2016, February 2017 - March 2017 (Totally 369 days)
Institute for Cognitive Systems (Prof. Gordon Cheng), Technical University of Munich, Munich, Germany; Visiting researcher
August 2014 - September 2014
University of California, Davis (U.C.Davis); Language School
Copyright @ Kuniyuki Takahashi All Rights Reserved.