Tomohiro Yokoyama
Papers
2
Total Citations
7
H-Index
2
About
Tomohiro Yokoyama is a researcher whose work bridges the fields of robotics and artificial intelligence, with a particular focus on reinforcement learning and computer vision. His most notable contribution is the pioneering application of Q-Learning to enable a humanoid robot to autonomously acquire the complex giant-swing motion on a horizontal bar—a task traditionally reliant on predefined trajectory planning. This 2010 study, which has garnered 4 citations, demonstrated that robots could learn dynamic, acrobatic behaviors through trial and error, offering a novel paradigm for adaptive control in sports robotics. More recently, Yokoyama has explored the use of convolutional neural networks (CNNs) for texture classification, specifically applied to identifying surface conditions on power lines. This 2019 work, with 3 citations, showcases his versatility in applying deep learning to practical infrastructure inspection challenges. While his citation counts are modest, Yokoyama’s early integration of reinforcement learning for complex motor skills represents a forward-thinking approach that predates the widespread adoption of such techniques in robotics, marking him as an innovator in learning-based control systems.
Research Focus
Key Achievements
Top Papers
- 1Realization and analysis of giant-swing motion using Q-Learning4 citations · 2010
- 2