Paper 'Is Weakly-supervised Action Segmentation Ready For Human-Robot Interaction? No, Let’s Improve It With Action-union Learning' selected as IROS Best Paper Award Finalist on Cognitive Robotics; Developed the Fujitsu Gymnastic Judging Support System, which was successfully applied in 10 apparatuses at the 52nd FIG Artistic Gymnastics World Championships, and applied for multiple patents.
Research Experience
Responsible for developing the multi-camera gymnast detection and tracking modules for the Fujitsu Gymnastic Judging Support System. Published related techniques in multiple international conferences and journals.
Education
Bachelor's degree from Nanjing University (China); Master's and Doctoral degrees from Nara Institute of Science and Technology (Japan).
Background
Currently working as a senior researcher at Fujitsu Research, with research interests including multi-camera motion capturing, pose estimation, and action recognition.