Electrical and Computer Engineering
Seminar
AI Vision-Enabled Pediatric Prosthetic Hand: A Low-Cost, Intelligent Solution for Children
Md Abdul Baset Sarker
PhD Candidate
Clarkson University, Potsdam, NY 13699, USA
Abstract: Children born with upper limb deficiencies often struggle with conventional prosthetic solutions, especially EEG- or EMG-based systems, which require brain or muscle signals and training, making adaptation difficult. AI vision-based systems offer an alternative to address this challenge. AI vision-enabled pediatric prosthetic hand is designed for children aged 10-12. The project focuses on developing a low-power, FPGA-based system with a camera for object detection and grasping, along with a soft structure suitable for children. Key features include: a) A wrist-mounted camera for artificial sensing in various hand tasks, b) Real-time object detection and distance estimation for grasping, and c) Low-power operation to function within resource constraints.
Today’s session will demonstrate a low-power FPGA-based system with optimized deep-learning models for real-time object detection in pediatric prosthetic hands. We will highlight how quantization and pruning techniques are used to run efficiently on resource-constrained edge devices. Key topics include multi-sensor systems for adaptive grasping, lightweight model deployment in the deep-learning processor unit (DPU), and balancing computational efficiency with accuracy for affordable, child-friendly prosthetics.
Co-sponsored by: HKN-Gamma Gamma
Speaker(s): Abdul Baset Sarker
Room: 194, Bldg: CAMP, Potsdam, New York, United States, Virtual: https://events.vtools.ieee.org/m/473920