[Biography]

Youngki Lee is an associate professor in the Computer Science and Engineering Department of Seoul National University. He received his PhD in 2012 from the computer science department in KAIST, Korea, and worked at Singapore Management University till 2018. He has broad research interests in building experimental and creative software systems, which involve multi-dimensional considerations across operating systems, applications, and users. More specifically, he has been developing ML/DL-driven mobile and sensor systems to enable always-available and highly enriched awareness of human behavior, emotion, and surrounding contexts. Also, he addressed core technical challenges such as improving energy efficiency, handling concurrency, and improving recognition accuracy. Furthermore, he has been building and deploying innovative mobile sensing applications in various application domains such as daily healthcare, childcare, education in collaboration with domain experts. He published a stream of his work in top-tier conferences and journals such as ACM MobiSys, ACM MobiCom, ACM SenSys, ACM UbiComp, IEEE TMC. He has also been serving technical program committees and organizing committees of such conferences including program co-chairs for UbiComp 2018. More details about him are available at http://youngkilee.blogspot.com.

[Abstract]

In this talk, I will speak about our on-going research efforts to build mobile deep learning systems for emerging extended reality applications. Specifically, I will introduce two specific works: the AR-based person identification system for crowded urban spaces and the mobile GPU coordination system to handle concurrent neural network workloads. I will share the challenges in building these systems and our solution ideas to address the problems.

ZOOM : https://zoom.us/j/8978217407?pwd=d1pOZmF1OWlseEdZRVBpV3VuSkl3dz09

ID : 897 821 7407

PW : 1nTQDY