Papers
27
Total Citations
292
H-Index
8
About
Naoki Wake is a robotics and artificial intelligence researcher whose work sits at the cutting edge of human-robot interaction, language-driven robot control, and learning-from-observation systems. His most celebrated contributions involve harnessing large language and vision-language models — including ChatGPT and GPT-4V(ision) — to bridge natural human communication and autonomous robot execution. His 2023 paper on ChatGPT-empowered long-step robot control (92 citations) demonstrated how few-shot prompting could translate everyday language instructions into executable robotic actions, while his 2024 work on GPT-4V(ision) for robotics (60 citations) extended this vision by enabling robots to learn manipulation tasks directly from human video demonstrations. Beyond foundation model integration, Wake has made meaningful advances in learning-from-observation frameworks, task-grasping, multimodal robot teaching, and gesture generation for social robots. His earlier work also tackled practical challenges such as ego-noise reduction for humanoid speech interfaces. A contributor to the emerging discourse on holistic Agent AI, Wake consistently pursues systems that unite perception, language understanding, and physical action. With nearly 200 cumulative citations across a focused body of work, he is establishing himself as a significant voice in next-generation intelligent robotics.
Research Focus
Key Achievements
Top Papers
- 1
- 2
- 3
- 4
- 5Task-grasping from a demonstrated human strategy11 citations · 2022
- 6
- 7Position Paper: Agent AI Towards a Holistic Intelligence9 citations · 2024
- 8Deep Gesture Generation for Social Robots Using Type-Specific Libraries8 citations · 2022
- 9Grasp-type Recognition Leveraging Object Affordance8 citations · 2020
- 10