CV

Profile

First-year PhD student at Science Tokyo in Shinoda Lab, working on robotic manipulation using Vision-Language-Action (VLA) models. My broader interests include robotics, reinforcement learning, foundation models, and explainable AI.

Education

• PhD, Shinoda Lab, Science Tokyo, Tokyo, Japan, 2026-present
  • Research topic: robotic manipulation with Vision-Language-Action models
• Master’s Degree in the SMILAB lab, Keio University, Tokyo, Japan, 2023-2025
  • Advisor: Prof. Komei Sugiura
  • Research topics: explainable AI, reinforcement learning, robotics
• Engineering Degree, Ecole Centrale de Nantes, Nantes, France, 2021-2025
  • Specialization: Artificial Intelligence
  • GPA: 3.7 (Year 1), 3.83 (Year 2)
• ‘Classes préparatoires’, Lycée Gustave Eiffel, Bordeaux, France, 2019-2021
  • A 2-Year intensive program preparing for the national competitive exams for entry to French schools of engineering.

Research and work experience

• 2026-present: PhD Research, Shinoda Lab, Science Tokyo
  • Working on robotic manipulation using Vision-Language-Action models.
  • Studying foundation-model-based approaches for embodied robotic agents.
• 2023-2025: Master’s Research, SMILAB, Keio University
  • Developed visual explanation methods based on Layer-Wise Relevance Propagation for ResNet.
  • Worked on reinforcement learning and robotics-oriented policy learning.
  • Published work at JSAI 2024 and ECCV 2024.
• April 2023 - September 2023 : Machine Learning Intern
  • R&D department at Aubay
  • Developed a web application for automatic meeting summarization using Python, Flask, and BERT models.
  • Deployed the application on AWS with Docker.
  • Collaborated with 6 team members using Agile methodology.
  • Presented the project to more than 200 company employees.

Skills

• Languages
  • French : Native
  • English : Fluent
  • Japanese : Intermediate
  • Spanish : Notions
• Deep Learning with PyTorch and JAX
  • Explainable AI
  • NLP
  • Robotics
  • Reinforcement Learning
  • Image Recognition
• Other
  • Unity
  • SQL
  • C++
  • LaTeX
  • Git/GitHub

Publications

Layer-Wise Relevance Propagation with Conservation Property for ResNet

Otsuki, Seitaro, Tsumugi Iida, Félix Doublet, Tsubasa Hirakawa, Takayoshi Yamashita, Hironobu Fujiyoshi, and Komei Sugiura. "Layer-Wise Relevance Propagation with Conservation Property for ResNet." In European Conference on Computer Vision, pp. 349-364. Springer, Cham, 2025.

C3-LRP: Visual Explanation Generation based on Layer-Wise Relevance Propagation for ResNet

Félix DOUBLET, Seitaro OTSUKI, Iida TSUMUGI, and Komei SUGIURA. "C3-LRP: Visual Explanation Generation based on Layer-Wise Relevance Propagation for ResNet." 人工知能学会全国大会論文集, vol. JSAI2024, pp. 4Q1IS2c03-4Q1IS2c03, 2024.

Talks

Presentation at JSAI24 of my paper C3-LRP: Visual Explanation Generation based on Layer-Wise Relevance Propagation for ResNet

April 24, 2024

Talk at JSAI24, Hamamatsu, Japan

My hobbies

• I played tennis competitively for 12 years and still play from time to time
• Avid anime enthusiast with over 100 titles watched
• I was the treasurer of the journal club at Centrale Nantes