Zero-Shot Generation of Semantic Co-Speech Gestures -- new paper out in ACM TiiS
A new paper is just out in ACM Transactions on Interactive Intelligent Systems: “ImaGGen: Zero-Shot Generation of Co-Speech Semantic Gestures …
Welcome to the research group for Social Cognitive Systems of the Faculty of Technology and the Center for Cognitive Interaction Technology (CITEC) at Bielefeld University, headed by Prof. Dr.-Ing. Stefan Kopp.
Modern A.I. systems excel at many tasks but lack essential capabilities for interacting and cooperating with humans and other agents. We investigate how A.I. systems can attain the level of cognitive, communicative and cooperative abilities that is required for successful (proto-)social interaction, human-agent collaboration, and even teamwork. Our research joins methods from A.I., Cognitive Science and Human-Agent/Robot Interaction. We study human cooperative and communicative behavior in order to develop A.I. models and systems with capabilities such as online Theory of Mind, social signal understanding, predicting others, planning joint actions, or communicating with language and other modalities in situated dialog. We bring them to bear in socially assistive robots, embodied conversational agents or collaborative A.I. agents that enable humans and agents to interact and cooperate effectively in educational, workplace or home scenarios.
A new paper is just out in ACM Transactions on Interactive Intelligent Systems: “ImaGGen: Zero-Shot Generation of Co-Speech Semantic Gestures …
Stefan delivered a keynote to the Symposium on Multimodal Communication (MMSYM 2026) held at KU Leuven, Belgium, discussing the trends, challenges and …
New papers got accepted and will be presented soon at the following conferences: “Natural Language Question Understanding” with Christoph …
After 5 years of research, Hendric Voss defended his PhD thesis on generating semantically meaningful and contextually relevant gestures – …
Annika Österdiekhoff successfully defended her PhD thesis on modeling a sense of control in RL agents. Congratulations!! And many thanks to Verena …
Excited to share that we are part of a new MSCA Doctoral Network on “Embodied Multimodal Human Communication and Modeling” (EMHCAM) that …
A new paper “Giving AI agents a sense of control facilitates reinforcement learning in multitasking scenarios” is now out and publicly …
Join us at the ZiF in Bielefeld for a workshop on “Multimodal creativity in speech and gesture production” on Dec 1st and 2nd! We are …