CEID Seminar & Social Hour: "Printed Intelligence: Machine Learning Classification with Printed Electronics", Γιώργος Ζερβάκης, Επίκουρος Καθηγητής, CEID

Εικόνα stratis
Σας ενημερώνουμε για την παρακάτω ομιλία η οποία θα δοθεί στα πλαίσια της σειράς εκδηλώσεων "Σεμινάριο CEID & Social Hour" και των ΔΠΜΣ ΥΔΑ, ΣΜΗΝ και ΟΣΥΛ.
CEID Seminar & Social Hour: "Printed Intelligence: Machine Learning Classification with Printed Electronics", Γιώργος Ζερβάκης, Επίκουρος Καθηγητής, CEID
Τίτλος:  Printed Intelligence: Machine Learning Classification with Printed Electronics
Ομιλητής: Γιώργος Ζερβάκης, Επίκουρος Καθηγητής, CEID
Ημερομηνία-χώρος: Παρασκευή 31 Μαρτίου,  3-5μμ, ΤΜΗΥΠ, αμφιθέατρο Γ
Περίληψη: Printed electronics appears as a viable solution to bring computing and intelligence in domains such as disposables (e.g., packaged foods, beverages), smart packaging, low-end healthcare products (e.g., as smart bandages), in-situ monitoring, as well as the 10-trillion market of fast moving consumer goods (FMCG). Printed electronics is increasingly recognized as a key enabler for the Internet of Things as part of the “Fourth Industrial Revolution”, whose core technology advances are functionality and low-cost. While the impact of computing is almost ubiquitous, such application domains have not seen considerable penetration of computing that could help, for example, with identification and tracking, quality monitoring, brand authentication, or interactivity. Such domains pose requirements for ultra-low cost (even sub-cent) and conformality that lithography-based silicon technologies cannot satisfy. Silicon systems cannot meet stretchability, porosity, and flexibility requirements, while the high manufacturing, packaging, and assembly costs of silicon prevent sub-cent cost. Despite the attractive features, printed electronics technology exhibits several prevalent limitations. The large feature sizes lead to high device latencies and low integration density, i.e., orders of magnitude lower than in silicon, hindering the realization of complex systems such as machine learning (ML). The fact that in printed applications (healthcare and medical devices, smart packaging and logistics, sensors for IoT, industry and environmental monitoring, and automotive) the core task is mainly classification highlights the high need for realizing printed ML circuits albeit the inherent physical limitations. This talk will present our research activities towards enabling, for the first time, printed ML classification circuits by exploiting, among others, non-conventional computing approaches.
Σχετικά με τον ομιλητή: Georgios Zervakis received the Diploma and Ph.D. degrees from the School of Electrical and Computer Engineering (ECE), National Technical University of Athens (NTUA), Greece, in 2012 and 2018, respectively. He is currently an Assistant Professor at the Department of Computer Engineering & Informatics Department, University of Patras, Greece. From 2019 to 2022 he was a Research Group Leader at the Chair for Embedded Systems (CES), at the Karlsruhe Institute of Technology (KIT). Dr. Zervakis has worked in many EU-funded research projects as a research associate of the Institute of Communication and Computer Systems (ICCS), Athens, Greece. He serves as a reviewer in many IEEE and ACM Transactions journals and is also a technical program committee member of the major design conferences. His main research interests include low-power design, accelerator microarchitectures, approximate computing, design automation, printed electronics, and machine learning.

Startup Growth Lite is a free theme, contributed to the Drupal Community by More than Themes.