|
|
|
|
Research Descriptions |
|
My research goal is to realize Intermittent Artificial of Things (iAIoT), enabling battery-less IoT devices to intermittently execute deep neural networks (DNN) via ambient power. iAIoT is a novel research direction at the intersection of intermittent computing and deep learning, and once realized, would create innovative applications.
My research team has released a suite of system runtime and libraries, facilitating AI and IoT application developers to easily build low cost, intermittent-aware inference systems. In particular, an intermittent operating system (TCAD’20), which was the first attempt to allow multitasking and task concurrency on intermittent systems, makes complicated intermittent applications increasingly possible. The HAWAII middleware (TCAD’20), which comprises an inference engine and API library, enables hardware accelerated intermittent DNN inference. In addition, the iNAS framework (TECS’21) was the first framework that introduces intermittent execution behavior into neural architecture search to automatically find intermittently-executable DNN models. HAWAII and iNAS received the Best Paper Awards, respectively, for two years in a row at IEEE/ACM CODES+ISSS 2020 and 2021.
|
|
|
|
|
|
|
|
|
|
|