Host: The Japan Society of Mechanical Engineers
Name : [in Japanese]
Date : June 08, 2016 - June 11, 2016
This paper proposes a novel control method which incorporates an EMG pattern classifier and a vision-based object classifier to control various motions of an electric artificial arm. The deep convolutional neural network is adopted as the object classifier and the posture of the electric artificial arm is controlled based on its classification result.The EMG signals are also used for controlling the phase of motion. To verify the proposed control method, a validation experiment was executed with 22 target objects. The 55 images for each target object were collected on the Web. The result revealed that the proposed method has high potential to control the various motions of the electric artificial arm.
