In the realm of artificial intelligence-based medical decision support systems, achieving the delicate balance in trust calibration becomes paramount — navigating the fine line between overtrust and undertrust. This presentation, titled "Heartbeat of Confidence: Calibration of Trust in Medical AI," explores the nuanced dynamics of trust in AI systems, unveiling strategies to modulate trust levels effectively. The discussion focuses on the role of explainable AI, acknowledging its potential to increase trust while addressing the dilemma of fostering understanding versus creating false hope. Additionally, the often underestimated practice of validating AI models on truly external data takes center stage, offering fresh perspectives on its significance and proposing approaches for increasing external trust in validation efforts.

Vortragender

Prof. Dr. Marc Aubreville, TH Ingolstadt

Zeit und Raum

Mittwoch, 25.10.2023 um 17:00 Uhr im K001

Abstract

Connected cars have ushered in a new era of automotive technology, seamlessly integrating vehicles with the digital world. These smart cars come equipped with various applications that enhance both the driving experience and overall vehicle functionality. Data connectivity applications in connected vehicles have revolutionized the way we experience driving. By harnessing the capabilities of the internet, these applications enable a wide array of features that enhance safety, convenience, and overall driving efficiency. Through dedicated applications, car owners can remotely lock/unlock doors, start the engine, or even adjust climate control settings. This level of control improves security and convenience, especially in situations where a remote start on a cold morning or unlocking the car for a family member becomes necessary. Furthermore, data connectivity applications contribute significantly to driver and passenger safety. Emergency services can be automatically alerted in the event of a crash, ensuring swift response times and potentially saving lives. By leveraging real-time data, these applications enhance various aspects of driving, making it safer, more efficient, and incredibly convenient for users.

Vortragende

Katharina Pletz, Solution Architect, Cloud Infrastructure Services, Capgemini

Zeit und Raum

Montag, 06.11.2023 um 18:00 Uhr, Raum K002
Link zum Zoom-Meeting: https://oth-regensburg.zoom.us/meeting/register/tZEtdO6sqD0iE9c2ffotxDQU5Fx-SvcvcIBZ

 

Abstrakt

1. Konkrete Schritte im Hinblick auf die Cyber Kill Chain eines Angriffs diskutieren
2. Möglichkeiten der Gegenwehr sowie präventive Maßnahmen herausstellen
3. Desaster Recovery Strategien im Hinblick auf Anlaufzeiten für die Produktion vergleichen
4. Nachgelagerte Analysemöglichkeiten bezüglich des Angriffs vorstellen
    

Vortragender

Sebastian Pfaller, Head of Product Management bei ondeso GmbH

Zeit und Raum

Dienstag, 14. November um 18:00 Uhr im K002

Abstract

Sebastian Erdenreich wird anhand von Beispielen aufzeigen, wie sich die Generative AI Modelle in Software- projekten anwenden, integrieren und erweitern lassen. Zu- dem werden auch Ansätze und Frameworks vorgestellt, die bisher weniger bekannt, aber mindestens genauso nützlich sind (Embeddings, Vektordatenbanken). Der Vortrag richtet sich an alle Studierenden, die sich für ChatGPT und Co. in- teressieren und über (software-)technische Grundkenntnisse verfügen.

von

Dr. Sebastian Erdenreich (trinnovative GmbH)

Termin

Mittwoch, 22. November 2023

Zeit

17:30 Uhr

Ort

Galgenbergstr. 32, K001

Abstract

Laut dem Gartner Hype Cycle befinden sich viele Ausprägungen von AI gerade am "Peak of Inflated Expectations". Viele Mittelständler und auch kleinere Unternehmen sehen sich dabei mit einer Flut an neuen Anwendungen konfrontiert.
Einer dieser Mittelständler ist die HORSCH Maschinen GmbH. Sie sieht ihr Kerngeschäft in der Herstellung von landwirtschaftlichen Maschinen, nichtsdestotrotz werden kontinuierlich Möglichkeiten zur Digitalisierung / Automatisierung evaluiert.

Maximilian Braun wird an einem Projekt darstellen, wie KI als Katalysator für eine erfolgreiche Anwendung wirken kann. Dabei werden die Phasen der Entwicklung beleuchtet und die Frage beantwortet: "Wie viel KI braucht der Mittelstand wirklich?"

von

Dr. Maximilian Braun (Horsch Maschinen GmbH)

Termin

Mittwoch, 13. Dezember 2023

Zeit

18:00 Uhr

Ort

Galgenbergstr. 32, K018

Ansprechpartner

Prof. Dr. Brijnesh Jain

Abstract

This presentation will delve into the dynamic interplay between Smart City services and citizen engagement and how true success of Smart Cities goes beyond the implementation of digital solutions it hinges on the active involvement of and empowerment of citizens in shaping the cities that they call home. Mr. Sarkar will unravel the profound impact of citizen participation on the trajectory of Smart Cities development by exploring five key dimensions:

1. Citizen-centric Urban Planning
2. Digital Democracy and Inclusive Governance
3. Data-driven Insights for Responsive Cities
4. Navigating Challenges and Ethical Considerations
5. Cultivating Sustainable Futures

 

Speaker

Amitrajit Sarkar - Senior Academic Member of Ara Institute of Canterbury, Te Pukenga, New Zealand

Date

Thursday, 14.12.2023

Start

18:30 Uhr in lecture hall K001 or via Zoom

Organizer

Prof. Dr. Markus Westner

Abstract

In today's volatile and rapidly changing business environments, technology has a fundamental role to play in sustainable business success. Meeting the challenge of the increasing technology complexity in business requires efficient technology governance, organizational adaptations and new approaches to steer IT. Using case studies of real project challenges and inventive IT steering, the talk will provide insights and solutions for mastering technology through steering mechanisms in daily business and projects life.

Speaker

Philipp Meissner (Manager, CapGemini Invent), Alina Reher (Senior Consultant, Capgemini Invent), Tobias Held (Consultant, CapGemini Invent)

Date

19.12.2023 @ 3:30pm

Room

K016

Abstract

BSH Hausgeräte GmbH is the largest manufacturer of home appliances in Europe and one of the leading companies in the sector worldwide. In this
talk, after a brief introduction of our company and our location in Regensburg, we will provide an overview of the current status of Voice
Interaction (and Spoken Dialogue Systems) for our global home appliance portfolio. We will present an overview of the different technical
solutions (on-the-edge, cloud-based, and hybrid) available today and give you a sneak preview of prototypes that we are currently working on,
including a first system that harnesses Large Language Models to demonstrate our vision of the next step towards an enhanced User Experience.

Vortragende

Lilly Heindl, Dr. Manuel Kirschner - Business Excellence Officer, Speech Technology Specialist, BSH Hausgeräte GmbH, Regensburg

Datum

Mittwoch, 10.01.2024, 15:30 Uhr in Raum K002

Abstract

Quantum computing emerged as a promising approach to cope with NP-hard problems. However, as the number of qubits increases, their fault tolerance decreases exponentially, characterizing the so-called Noisy Intermediate-Scale Quantum (NISQ) era. Moreover, the number of qubits available in quantum computers is far from the ideal.

This talk reviews fundamental concepts about quantum machine learning and two applications: question-answering in natural language processing and pattern classification using the Optimum-Path Forest.

Meet Professor Papa at a small reception in front of the lecture hall with beer and pretzels!

Speaker

Prof. Dr. João Paulo Papa, São Paulo State University, Brazil

Date

Mittwoch, 21.02.2024, 16 Uhr

Room

Raum K 001

More information to Prof. Papa

Prof. Dr. João Paulo Papa is Humboldt fellow and guest for three months at the ReMIC lab (Regensburg Medical Image Computing, Head: Prof. Dr. Palm) at OTH Regensburg as the host institution.

Prof. Papa is full professor at the São Paulo State University (Unesp) in Brazil and has a strong background in Machine Learning.

He finished his PhD in Computer Science in 2008 with a work about Supervised Classification of Patterns using a Forest of Optimal Paths at State University of Campinas (UNICAMP), Brazil. This was followed by Post-Doc positons at UNICAMP, Brasil and Harvard University, USA. Since 2016, he was appointed as professor at Unsesp, Brazil and head of the Recogna lab. In 2017 he received his first Capes/Humboldt fellowship visiting OTH Regensburg and the ReMIC lab for several months. Since 2017 he is president of the Brazilian Chapter of the International Assiciation of Pattern Recognition and since 2020, he is Chair of the IEEE Task Force for Business Intelligence and Knowledge Management.

He published a huge amount of scientific papers (h-Index: 59) with more than 12k citations. He received several awards like the Future Startup Award (2022), Honorable Mention at the Workshop Tests and Dissertations at Sibgrapi (2022), the Inventor Prize by the Innovation Agency at University of Campinas (2019), the Best Paper Award at the 2nd International Conference on Biosignal Analysis, Processing and Systems of the IEEE Signal Processing Society (2018) and the Best Paper published in Pattern Recognition (2014), to mention only a few of them.

Get more insights into his impressive work at Google Scholar!