| 日時(Date) |
2026年01月23日 (金) / Jan. 23rd, 2026 (Fri.) 4限-6限 (15:00--18:00) / 4th-6th period (15:00--18:00) |
|---|---|
| 場所(Location) | Panasonic IS 大講義室[L1]（IS） |
| 司会(Chair) | Monica sensei |
| 講演者(Presenter) | TALK-1: Prof. Dr. Elisabeth André (University of Augsburg, Germany) |
| TALK-2: Dr. Patrick Gebhard (German Research Center for Artificial Intelligence, Germany) | |
| TALK-3: Prof. Dr. Thomas Rist (Technical University of Augsburg, Germany) | |
| 題目(Title) | TALK-1: Sign, Speak and Socialize: AI-Based Tools for Expression, Inclusion |
| TALK-2: Framework for Artificial Empathic Technology-Operationalization of Empathy for HCI | |
| TALK-3: Empowerment and Leveraging AI-Assistants in Computer Games and for Game Making | |
| 概要(Abstract) | TALK-1:Advances in analytic and generative AI present promising opportunities for enhancing communication by augmenting individuals' perceptual and expressive capabilities. This talk will explore recent advancements and practical applications of conversation-enhancing technologies, with a focus on their potential benefits for individuals facing communication challenges. Collaborative efforts have resulted in innovative applications tailored to diverse user needs. Examples include signing avatars designed to assist the hearing impaired, as well as more expressive voices for users of AAC (Augmentative and Alternative Communication) devices. Additionally, virtual reality environments have been developed to facilitate social skill training for both children and adults, leveraging role play interactions with virtual agents. We will discuss how analytic and generative AI can serve as powerful tools to overcome communication barriers, foster authentic expression, and promote social inclusion. By analyzing the potential of these technologies to support users with varying communication abilities, we aim to address the challenges inherent in designing empathetic solutions that cater to the specific needs of diverse audiences, while mitigating the risk of reinforcing stereotypes. |
| TALK-2:The talk introduces a framework for artificial empathic technology. It models empathy as a consistent processing and action principle: from signals to analysis and interpretation to the regulation of emotional states and regulative social interaction dynamics to appropriate expression. The starting point is computer model-based methods for emotion recognition, along with more detailed computational modeling of emotions; the underlying concepts of emotion and empathy are critically discussed. Examples from current and previous projects illustrate the building blocks and highlight areas where further research is needed. A compact evaluation framework (explainability, robustness, fairness, interaction quality) serves as a checklist for research and prototyping. | |
| TALK-3:The application of AI in computer games holds significant potential to enhance gameplay and to support developers in the game making process. In a first showcase, we present iVoice, which is a simulation of a player's “inner voice” while exploring an interactive 3D-environment. The player’s actions and anticipated observations in the game world as well as the course of game play form the basis for the generation of inner voice utterances. In a gaming context, using an inner voice opens up various possibilities. Firstly, it can be parameterized with regard to its linguistic realization, including, among other things, message length, language style and tone, word choices, lexical frequency of words, or the use of certain rhetorical figures. Secondly, when triggered by a game world event, the message content can be biased to influence the player’s decision-making. Thirdly, an inner voice can provide clues about the background story and steer the player's construction of a self-narrative. Another strand of research starts from the observation that map-like representations, such as level-floor plans, play a pivotal role in level design by capturing essential spatial aspects like topology and navigable areas. In a series of experiments, we were prompting multi-purpose AI assistants with textual descriptions of tasks related to level-design alongside map-like representations of the spatial context. Our observations lead us to conclude that AI assistants capable of multimodal understanding (text and visual) and spatial reasoning can effectively assist in such tasks. This is particularly viable in an interactive setting where humans can guide the process by commenting on and requesting refinements to the output in follow-up prompt. The main potential of such AI-assistants lies in their ability to understand a user’s intent and then work out a solution. Level designers can simply use their own words and visual sketches to instruct an AI assistant. This is particularly helpful as level designers are not necessarily experts in computational geometry. Furthermore, AI-assistants can explain their reasoning, and users can critique proposals and request corrections and refinements. | |
| 講演言語(Language) | English |
| 講演者紹介(Introduction of Lecturer) | Elisabeth Andre is a full professor of Computer Science and Founding Chair of HumanCentered Artificial Intelligence (AI) at Augsburg University in Germany. She has a long track record in multimodal human-machine interaction and AI-based human behavior analysis with applications in social coaching, psychotherapy and mental health diagnosis. Her work has won many awards including the Gottfried Wilhelm Leibnitz Prize 2021 of the German Research Foundation (DFG), the most prestigious German research award. In 2017, she was elected to the CHI Academy, an honorary group of leaders in the field of Human-Computer Interaction. To honor her achievements in bringing Artificial Intelligence techniques to Human-Computer Interaction, she was awarded a EurAI fellowship (European Coordinating Committee for Artificial Intelligence) in 2013. In 2019, she was named one of the 10 most influential figures in the history of AI in Germany by National Society for Informatics (GI). Elisabeth André is a member of the Bavarian Academy of Sciences and Humanities, the Academy of Europe, the Asia-Pacific Artificial Intelligence Association, the National Academy of Science and Engineering acatech and the German Academy of Sciences Leopoldina. Furthermore, she is a fellow of the European Laboratory for Learning and Intelligent Systems (ELLIS). |
| Patrick Gebhard is a DFKI Research Fellow and Principal Researcher in the department for Cognitive Assistants at the German Research Center for Artificial Intelligence (DFKI) in Saarbrücken, where he has led the Affective Computing Group since 2007. His work centers on creating computational models of affect, emotion regulation, and social behavior, as well as on researching the use of socially interactive agents. Notably, he created the ALMA model of emotions, moods, and personality, and has made significant contributions to mechanisms of empathy and trust in human–AI interaction. Gebhard and his group have created frameworks and tools, such as Visual SceneMaker, which are widely used to build and evaluate systems with socially interactive agents. These tools and computational models of affect have been applied in numerous national and EU projects across healthcare, education, and HRI, where this collaborative approach has been instrumental in enhancing trustworthiness and effectiveness. His research centers on advancing computational models of affect by incorporating regulatory processes and social interaction. These models represent an important foundation for future interactive assistance systems with human-like emotional intelligence in therapy, education, and work. Lately, his research has incorporated the use of large vision and language models for deeper modelling of internal and external context. | |
| Prof. Thomas Rist studied computer science at the University of the Saarland, Germany, and earned a PhD in 1996 from the same university. From 1987 - 2004 he was researcher, senior researcher, and research fellow at DFKI, the German Research Center for Artificial Intelligence. He has a long track record in the coordination of applied research projects funded by the European Union, national German funding agencies as well as industries. In 2004 he was appointed a professor of Computer Science at Technical University of Augsburg, Germany. He was a driving force in the development and launch of the two interdisciplinary study programs on Interactive Media (BA/BSc) and Interactive Media Systems (MA). Currently he contributes to both study programs with courses on the technical conception of computer games, games programming, AI for computer games, and UI/UX evaluation methods for games. In addition, he supervises various games projects (for examples see https://showcase.tha.de/ia-ims?). His current research activities include approaches to leverage LLM-based AI in computer games |