A quantitative review of automated neural search and on-device learning for tiny devices

This paper presents the state-of-the-art review of the different approaches for Neural Architecture Search targeting resource constrained devices such as microcontrollers. As well as the implementations of On-Device learning techniques for those devices.