Yoshiki Obinata
Papers
11
Total Citations
96
H-Index
6
About
Yoshiki Obinata is a pioneering researcher at the intersection of robotics, computer vision, and natural language processing, with a focus on enabling robots to understand and interact with their environments through vision-language models (VLMs). His major contributions center on developing state recognition systems that allow robots to perceive continuous object changes—such as food cooking states—using pre-trained VLMs combined with black-box optimization and genetic algorithms. This work has been instrumental in advancing cooking robots that can execute real-world recipes by integrating foundation models with PDDL planning. Obinata’s research also extends to semantic scene difference detection for mobile patrol robots and novel approaches to open-vocabulary navigation using omnidirectional cameras and multiple VLMs. His most cited paper, "VQA-based Robotic State Recognition Optimized with Genetic Algorithm" (2023, 22 citations), demonstrates the power of combining visual question answering with evolutionary optimization. Additionally, his work on automatic diary generation that captures joint human-robot experiences and emotion-based descriptions showcases his interest in human-robot interaction. With over 90 total citations and a rapidly growing publication record, Obinata is establishing himself as a leading voice in making robots more perceptive, adaptive, and capable of operating in unstructured, dynamic environments.
Research Focus
Key Achievements
Top Papers
- 1VQA-based Robotic State Recognition Optimized with Genetic Algorithm22 citations · 2023
- 2
- 3
- 4
- 5
- 6
- 7
- 8
- 9
- 10