A survey on deep learning based IoT sensing

PhD Qualifying Examination


Title: "A survey on deep learning based IoT sensing"

by

Mr. Hua KANG


Abstract:

In the age of the Internet of things (IoT), physical objects are equipped
with computing, sensing and communication capabilities to build a
sensor-rich and interconnected world and thus enable a wide range of IoT
applications. Ubiquitous IoT devices generate massive data which are time
and space dependent, configuration related, heterogenous and noisy going
beyond the capabilities of traditional inference and learning approaches.
At the same time, deep learning methods have shown promising results in
several areas including signal processing, natural language processing,
and image recognition. Applying advanced deep learning methods to IoT
sensing enables lots of novel applications and revolutionizes the complex
interactions of human and physical surroundings.

In this survey, we present a detailed overview of deep learning based IoT
sensing. We first introduce the background and motivation of deep learning
based IoT sensing. Then we give an introduction of sensing modalities and
deep learning models respectively. Next we review the published systems
categorized by their applications, including activity recognition, gesture
recognition, fall detection, vital signs monitoring, sleep monitoring,
authentication and localization. Finally, we conclude the survey and
discuss some possible future directions.


Date:                   Wednesday, 12 February 2020

Time:                   4:00pm - 6:00pm

Zoom Meeting:           https://hkust.zoom.us/j/423828046

Committee Members:      Prof. Qian Zhang (Supervisor)
                        Prof. Gary Chan (Chairperson)
                        Prof. Bo Li
                        Dr. Wei Wang


**** ALL are Welcome ****