Development of a ubiquitous learning system for dexterous hand operation

It is not easy to learn the dexterous finger movements of a skilled person due to flexibility and complexity. Therefore, a lot of training and effort is needed for a novice in order to acquire proficiency. The role model and feedback information of experienced persons is important for learning skilled movements. In this paper, we evaluate a ubiquitous learning system for dexterous hand operation using the finger motion capture data measured by the Hand-MoCap system. Through this system, a novice can learn under the guidance of the virtual 3D hand reconstructed using the master´s motion capture data. We quantitatively evaluate the effectiveness of the system.