14th edition AI lunch meeting (TU/e internal)

Date
Thursday November 28, 2019 from 12:30 PM to 2:00 PM
Location
TU/e
Address
Het Kranenveld 12
Co-organizer
Data Science
Building
Gaslab - Building 12

About the event

Many 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) lunch meetings where various researchers talk about their projects.

PROGRAM

12:30 Johan Lukkien (Dean M&CS) Introduction
12:45 Peter de With, Video Coding & Architecture (EE) Using multi-spectral data analysis for early cancer detection
13:05 Ioulia Ossokina, Urban Systems and Real Estate (BE) Effects of automated driving technologies for the economy and society: will cities grow or decline?
13:25 Aarnout Brombacher, Systemic Change (ID) Risk based data acquisition for healthcare
13:45   Wrap up

ABSTRACTS

Peter de With
With the further growth of sensor technologies, visual sensing is increasingly complemented by sensing beyond the visible spectrum. For example, infrared sensing already plays an important role in assessing human behavior for surveillance applications. For healthcare, the question arises whether non-visible spectral sensing may facilitate the early detection of cancer. This question is addressed in the view of existing cancer detection methods  that rely now on heavily computer vision and deep learning techniques. Two cases based in multi-spectral data analysis will be briefly discussed for digital pathology and early cancer analysis.

Ioulia Ossokina
This talk is based on big data research on possible economic and societal impacts of AI technologies. We show that automated driving can lead both, to growth and decline of cities. We simulate spatial effects of automated driving for the Netherlands using LUCA, the Dutch spatial general equilibrium model. Two components of automation are accounted for: (i) more productive time use during car trips; and (ii) fast and comfortable door-to-door automated public transit. We find that the car component results in population flight from cities, while the public transit component leads to population clustering in urban areas. A combination of the two may result in the population concentrating in the largest, most attractive cities, at the expense of smaller cities and non-urban regions. Welfare benefits due to automation are considerable, with up to 10% coming from population relocation and changes in land use.

Aarnout Brombacher
In the field of healthcare, the role of data is gaining considerable importance. In the recent past it was possible to analyze and diagnose patients in hospitals and clinics at the moment that problems became imminent and afterwards during treatment. Before intervention, and after treatment was ended no- or only very little- information was available. In the last, approximately, ten years this has radically changed. Although new technology can provide huge benefits for (future) patients both in early detection of potential problems as well as in mitigating the consequences the clear and obvious need to respect autonomy and dignity of (future) patients will definitely lead to discussions what technology to apply where, to whom and when. This presentation will present developments in the field of “localized AI” that could provide research directions towards solutions in the context of the above problem.

Organizer

Artificial Intelligence

We study foundations of AI for the present and the future. We design new AI methods, develop AI algorithms and tools with a view at expanding the reach of AI and its generalization abilities.

In particular, we study foundational issues of robustness, safety, fairness, trust, reliability, tractability, scalability, interpretability and explainability of AI.

Currently, DAI includes five research groups: Uncertainty in AI, Generative AI, Automated ML, Data Mining, and Databases.