Date
Tuesday March 15, 2022 from 12:00 PM to 1:00 PMLocation
Online - MS TeamsOrganizer
EAISIMany TU/e researchers are advancing or using Artificial Intelligence in their research projects. To support cross disciplinary learning and to strengthen the TU/e AI network, EAISI organizes a series of (internal) meetings during lunch time, where various researchers talk about their projects, followed by Q&A.
PROGRAM
12:00 | Shane Ó’Seasnáin | Program Director at EAISI | Introduction of the event and community |
12:05 | Tanir Ozcelebi | Associate Professor at Interconnected Resource-aware Intelligent Systems, Department Mathematics and Computer Science | Federated Self-Training for Semi-Supervised Audio Recognition |
| 12:20 | Manil Dev Gomony | Assistant Professor at Electronic Systems, Department Electrical Engineering Patty Stabile | Associate Professor at Low-Latency Inter-Connect Networks, Department Electrical Engineering Mathias Funk | Associate Professor at Future Everyday, Department Industrial Design | Manil Dev Gomony and Patty Stabile on their cutting edge researches, reported from an EE Lab with Mathias Funk |
12:40 | Ömür Arslan | Assistant Professor at Dynamics and Control, Department Mechanical Engineering | EdgeAI challenges of mobile perception and decision making for light-weight robot autonomy |
12:55 | Wrap-up |
Edge AI Research at TU/e
“Edge AI” brings the technologies of edge computing, the Internet of Things (IoT) and artificial intelligence together. It is an import crossroads from Data Science towards High-Tech Industry, Mobility and Healthcare. The main advantages of Edge AI are lower-latency, better privacy, reduced power-consumption, and increased robustness. It is widely seen as being a big opportunity for Europe to gain a lead over the US and China, by building on our strengths in semiconductors and telecommunications. Edge AI is a term used to describe the combination of edge computing and AI. In traditional AI, models are developed using cloud-based systems with training and validation sets of data. The resulting models are then applied across a network in situ. Edge AI is different to this approach in that the model is developed locally. This approach has cost and performance advantages in terms of power-use, privacy and responsiveness. The principle is that when intelligence resides at the edge, data can be processed effectively and derived information can be shared on a need-to-know basis with other devices in the system. Apart from applications powered by machine-learning, i.e., data-driven AI, Edge AI is also suited to work in hybrid AI deployments that combine or even merge model-driven with data-driven approaches for learning.