Efficient on Device Object Detection for Prosthetic Arm

This research paper presents a novel project aimed at enhancing the functionality of prosthetic arms through the integration of efficient on-device object detection. The project leverages a Raspberry Pi, equipped with a Mobile net computer vision model, and a camera embedded within the palm of the prosthetic arm. The primary objective is to enable real-time detection of everyday objects within the arm's field of view. Upon object recognition, the prosthetic arm performs either a grip or grasp action, enhancing the user's interaction with the environment. This paper outlines the technical details, implementation process, and the potential impact of this innovative system on the lives of individuals with limb loss, offering increased autonomy and improved quality of life