On-sensor TinyML Implementation - Advancing Neuroscience Through Wearable Devices

Explore a 56-minute technical talk that delves into the intersection of TinyML implementation and neuroscience through wearable devices. Learn how artificial intelligence and machine learning are revolutionizing neurological diagnostics and management, from sleep pattern analysis to stroke detection and epilepsy diagnosis. Discover the development of ultra-low-power System-on-Chip (SoC) solutions for next-generation neuro-wearables, focusing on power-efficient (1mW) ML-based patient-specific biomedical processors. Understand the challenges and solutions in on-sensor processor design, particularly for implementing machine learning algorithms in disease detection, with special emphasis on identifying negative emotion outbursts in autistic patients. Gain insights into energy-efficient digital processing techniques and future directions in wearable healthcare systems from Assistant Professor Muhammad Awais Bin Altaf of Lahore University of Management Sciences.