SFSU Engineering Enters Phase 2 of Neural-Machine Interface Development with SSUP
San Francisco State University (SFSU) School of Engineering has officially entered Phase 2 of its project with Sony’s Sensing Solutions University Program (SSUP). Building on the early achievements of “Neural2,” which developed a real-time neural-machine interface (NMI) for controlling bionic arms, SFSU is now advancing its research to improve the efficiency and reliability of deep learning models for NMIs. This new phase, “Efficient and Robust Deep Learning Models for Neural-Machine Interfaces with Spresense Microcontroller,” aims to create even more dependable systems for prosthetics, rehabilitation, virtual devices, and augmented reality.
During Phase 1, SFSU’s team reached a key achievement by implementing a high-density electromyography (HD-EMG)-based NMI using Sony's Spresense edge board. This system recognized hand gestures by processing muscle signals in real-time through optimized deep-learning models. Research assistant Peter Chudinov presented the development at SSUP’s US End-Year Review Meeting, showing how promising this technology can be.
Phase 2 will focus on refining these deep learning models to boost performance and scalability in real-world settings. Additionally, SFSU is integrating the Spresense technology into courses and capstone projects for the second time, such as the “On-Device Machine Learning” course. This gives students valuable experience with advanced machine-learning applications for edge devices.
Led by Dr. Zhuwei Qin and Dr. Xiaorong Zhang, the SFSU engineering team is excited to advance NMI technology that is both functional and accessible, opening up new possibilities in neural-machine interfaces that could make a lasting impact across industries.
