Low-Energy Physiologic Biomarker Machine Learning Inference on a Wearable Device with GAP9 RISC-V Processor

Watch a 21-minute conference talk from the tinyML Summit 2023 exploring Mayo Clinic's comprehensive framework for developing machine-learning physiologic biomarkers on wearable devices. Learn about the institution's approach to human subject testing for dataset creation, machine learning methods for biomarker extraction, and implementation using Greenwaves GAP9 RISC-V processor technology. Discover specific examples focused on regenerative physiologic signal autoencoders and their feasibility in low-energy wearable platforms. Gain insights into performance optimization techniques, including memory reduction, computation resource management, and hardware mapping strategies for the Greenwaves GAP9 processor in custom wearable prototypes. Presented by Christopher L. Felton, Development Engineer IV from Mayo Clinic SPPDG, this talk demonstrates the practical application of tinyML in healthcare monitoring devices.