Seminar: Towards Edge AI: Collaboration of Connected Devices
JinYi Yoon
Presidential Postdoctoral Fellow
Department of Computer Science
Virginia Tech
Friday, September 26, 2025
2:30 - 3:45 p.m.
Classroom Building, Room 260
Abstract
Artificial intelligence (AI) continues to spread across nearly every field, but AI systems mostly rely on remote cloud servers. As the number of users and requests grows, this leads to communication bottlenecks and thereby increased latency. Cloud AI also raises severe privacy concerns, since users must upload all raw private data to use the service. More importantly, cloud AI becomes unavailable during network disconnections or server outages. Fortunately, there are numerous edge devices, such as smartphones, laptops, Internet-of-Things (IoT), or sensors, located near to users. To this end, edge AI has emerged as a promising approach by deploying AI directly on edge devices, thus alleviating the reliance on clouds and providing more reliable, local processing. However, individual edge devices often lack sufficient data, computational power, and resource capability. Although shifting towards edge-side poses these challenges, at the same time, it also brings the advantage of having more usable devices, enabling a potential collaboration among edge devices. In this talk, we will explore collaborative learning paradigms of connected edge devices for on-device AI, including efficient AI, federated/split/distributed learning, and knowledge transfer approaches.
Biography
JinYi Yoon is a presidential postdoctoral fellow in the Department of Computer Science at Virginia Tech. She has received her B.S., M.S., and Ph.D. degrees in the Department of Computer Science and Engineering from Ewha Womans University, South Korea. Dr. Yoon has broad interests in the intersection of Edge Networks and AI.