Talk Abstract: We describe the development of a noninvasive neuromotor interface that allows for computer input using surface electromyography (sEMG). We developed a highly sensitive and robust hardware platform that is easily donned/doffed to sense sEMG at the wrist and transform intentional neuromotor commands into computer input. We paired this device with an infrastructure optimized to collect training data from thousands of consenting participants. This allowed us to develop generic sEMG neural network decoding models with performant out-of-the-box generalization across people (median performance for test users on a continuous navigation task: 0.5 target acquisitions/second; discrete gesture detection task: 0.9 gestures / second; handwriting task: 19.6 words per minute).

Speaker Biography: David Sussillo is a Senior Staff Research Scientist at Meta Reality Labs and an Adjunct Professor at Stanford University, where he holds affiliations with the Electrical Engineering Department and Wu Tsai Neurosciences Institute. He earned his PhD in Computational Neuroscience from Columbia University, where he studied under Larry Abbott and developed work on learning in chaotic recurrent neural networks. His research spans the intersection of neuroscience, machine learning, and dynamical systems, focusing on understanding neural computation through the lens of population dynamics. Before joining Meta, he was a Senior Research Scientist at Google Brain and completed postdoctoral training at Stanford University with Krishna Shenoy. Sussillo’s contributions include work on context-dependent computation in prefrontal cortex, brain-machine interfaces, and the development of novel methods for analyzing artificial neural networks.