Abstract: The proliferation of edge devices and the gigantic amount of data they generate make it no longer feasible to transmit all the data to the cloud for processing. Such constraints fuel the need to move the intelligence from the cloud to the edge where data reside. In this talk, I will present our works on how we bring the power of deep learning to edge devices to realize the vision of Artificial Intelligence of Things (AIoT).

 

This talk consists of two parts. The first part focuses on how we address some of the most fundamental problems that act as the key barriers of achieving the vision of AIoT. First, I will present our work on designing adaptive frameworks that empower AI-embedded edge devices to adapt to the inherently dynamic runtime resources to enable elastic edge AI. Second, I will present our work on distributed on-device training that significantly enhances the training efficiency without compromising the training quality. Third, I will talk about our work on developing automated machine learning (AutoML) techniques that provide a scalable solution to the device deluge challenge in AIoT. In the second part of this talk, I will present how we use AI as the core component to design AIoT systems for a broad range of problem domains. I will focus on one killer application of edge computing, and present an AI-enabled distributed edge system for low-latency high-throughput large-scale live video analytics.

Short Bio: