The Current State of TinyML – Opportunities, Challenges and the Road Ahead
TimeThursday, December 9th1:30pm - 2:00pm PST
Special Session (Research Track)
Hosted in Virtual Platform
DescriptionTiny machine learning (tinyML) is a fast-growing field of machine learning technologies and applications including algorithms, hardware, and software capable of performing on-device sensor data analytics at extremely low power, hence enabling a variety of always-on use-cases and targeting battery-operated devices. tinyML systems are slowly adopted for multiple commercial applications and new systems on the horizon, and at the same time, significant progress is being made on algorithms, networks, and models. Further, what was initially considered low power applications, is now mainstream and commercially available. There is therefore a growing momentum demonstrated by the technical progress, ecosystem development and the need for benchmarking and evaluation methodologies. In this talk, we present an overview of the current state of the art, while we also identify challenges and opportunities. We also provide our vision for the road ahead.