IKIDA
Interactive AI algorithms & cognitive models for human-AI interaction

IKIDA is a BMBF funded interdisciplinary research group researching in interactive AI algorithms and cognitive models for human-AI interaction.

Figure: Overview of IKIDA research topics
Figure: Overview of IKIDA research topics

IKIDA is an interdisciplinary junior research group and consists out of 6 young scientists from the research fields of machine learning, robotics and cognitive science.

The overall technical and scientific goal of IKIDA is the development of interactive, probabilistic AI algorithms, which benefit from direct interaction with users. Enabling such direct interaction between AI algorithms and users is also intended to lower application hurdles of AI solutions and can potentially increase their sustainability and acceptance. For the development of the AI algorithms we particularly plan to incorporate and develop new cognitive models for human-AI and human-robot interaction.

The associated industry partners Energy Robotics (Darmstadt), Franka Emika (Munich) and Porsche Motorsport (Weissach) will help to find and evaluate practical relevant use cases for the developed interactive AI methods, such as automated data classification and robot skill learning.

IKIDA is also supported by the AI initiative AI:DA and the Centre for Cognitive Science .

A short demonstration on Interactive Reinforcement Learning With Bayesian Fusion of Multimodal Advice (see: Trick, S., Herbert, F., Rothkopf, C., & Koert, D. (2022). Interactive Reinforcement Learning With Bayesian Fusion of Multimodal Advice. IEEE Robotics and Automation Letters, 7(3), 7558-7565. DOI PDF (opens in new tab))

A short demonstration on Adapting Object-Centric Probabilistic Movement Primitives with Residual Reinforcement Learning (see Carvalho, J., Koert, D., Daniv, M., Peters, J. (2022). Adapting Object-Centric Probabilistic Movement Primitives with Residual Reinforcement Learning. Proceedings of the International Conference on Humanoid Robots (HUMANOIDS) 2022 (accepted) PDF (opens in new tab))

A short demonstration on Learning from Unreliable Human Action Advice in Interactive Reinforcement Learning (see Scherf, L.; Turan, C., Koert, D. (2022). Learning from Unreliable Human Action Advice in Interactive Reinforcement Learning. 2022 IEEE-RAS 21st International Conference on Humanoid Robots (Humanoids). PDF (opens in new tab) )

January 2024 Our paper “Are you sure? – Multi-Modal Human Decision Uncertainty Detection in Human-Robot Interaction” got accepted at the ACM/IEEE International Conference on Human Robot Interaction (HRI) taking place March 11-15 in Boulder, Colorado. See Publications for details.

October 9, 2023 Lisa Scherf and Eya Chemangui present “Multi-Modal Human Uncertainty Detection in Human-AI Interation” at the Computational Cognition Conference 2023 (ComCo 2023) in Osnabrück.

August 18, 2023 Susanne Trick was featured on the Hessen schafft Wissen podcast on the topic of AI Assistance in Everyday Life without Usage Hurdles (in German)

July 6, 2023 IKIDA participates in The Hessian AICon and presents current projects results.

June 14, 2023 Our paper “Interactively Learning Behavior Trees from Imperfect Human Demonstrations“ got accepted for publication in Frontiers in Robotics and AI. See Publications for details.

June 02, 2023 Two papers “What Can I Help You With: Towards Task-Independent Detection of Intentions for Interaction in a Human-Robot Environment“ and ”I^3: Interactive Iterative Improvement for Few-Shot Action Segmentation“ got accepted at the IEEE RO-MAN 2023 taking place August 28-31 in Busan, Korea. See Publications for details.

May 31, 2023 Our paper “Interacting with Large Language Models: A Case Study on AI-Aided Brainstorming for Guesstimation Problems“ and extended abstract ”Interactively Learning Behavior Trees from Imperfect Human Demonstrations“ got accepted at the HHAI 2023 taking place June 26-30 in Munich, Germany. See Publications for details.

April 27, 2023 IKIDA researcher Susanne Trick was honoured by the Federal Ministry of Education and Research (BMBF) and the German Informatics Society (GI) as AI Newcomer of the Year. Congratulations!
Learn more

April 23, 2023 Vildan Salikutluk presents the IKIDA paper ”Situational Adaptive Autonomy in Human-AI Cooperation“ @ the CHI'23 workshop AutomationXP23 in Hamburg.
See Publications for details

October 2022 The IKIDA team got accepted 3 papers at 2022 IEEE-RAS International Conference on Humanoid Robots (Humanoids 2022), taking place November 28-30 in Ginowan, Okinawa, Japan. See Publications for details

July 13, 2022 The IKIDA team visited Accso – Accelerated Solutions GmbH, an enterprise active in software engineering and IT consulting in a broad range of sectors like finance, telecommunication, media, automotive industry, logistics and more. We discussed the potential of interactive AI for industry 4.0 applications and are exited to supervise a master's thesis together with Dr. Xenija Neufeld on the topic ”Detection of Unexpected Human Behavior in Human-Robot Interaction in Shared Workspaces".

July 2022 Our work on Interactive Reinforcement Learning with Bayesian Fusion of Multimodal Advice (opens in new tab) has been published in Robotics and Automation Letters.

March 2022 We are happy to announce that our paper Bayesian Classifier Fusion with an Explicit Model of Correlation (opens in new tab) has been presented at AISTATS Conference 2022.

January 1, 2022 IKIDA presented and discussed its ongoing research and intermediate results at the BMBF online conference “AI All-Hands-Meeting”.

November 2021 We are very happy that our work from Collaboration with Niyati Rawal on AI for Generating Robotic Facial Expressions got accepted for publication in Frontiers in Robotics and AI. ExGenNet: Learning to Generate Robotic Facial Expression using Facial Expression Recognition
N Rawal, D Koert, C Turan, K Kersting, J Peters, R Stock-Homburg Frontiers in Robotics and AI, 382

October 29, 2021 We visited our associated industry partner Energy Robotics. We thank Dr. Stefan Kohlbrecher for showing us impressive live demos of their robots Spot and EXR-1. We look very much forward to now start working together on the scenarios of Interactive Teaching and Human Aware behavior Adaptation for Inspection Robots. (see photos below )

October 2021 As announced in May, 1st Workshop on Empowering Interactive Robots by Learning Through Multimodal Feedback Channel was held on October 22nd, 2021 at the International Conference on Multimodal Interaction 2021 with three keynote speakers: Dr. Judith Holler, Dr. Georgia Chalvatzaki and Dr. Heni ben Amor.

October 2021 As of October 1, 2021, Dr. Dorothea Koert has been appointed leader of the independent junior research group IKIDA She has been leading the team since the start of the project in October 2020. Learn more (in German)

June 2021 Welcome Cigdem Turan! Cigdem joins our team as of June 1st as a postdoc. Her research focus will be facial behavior understanding, affective computing, affective human-robot Interaction, and artificial moral agents.

May 2021 We will present our work on Learning Probabilistic Movement Primitives and Sequential Behavior Trees from Non-Expert users at the IEEE-RAS International Conference on Humanoid Robots (July 19-21st, 2021)! Accepted Paper: “Guided Robot Skill Learning: A User-Study on Learning Probabilistic Movement Primitives with Non-Experts” Moritz Knaust, Dorothea Koert.

May 2021 We are excited to announce the 1st workshop on Empowering Interactive Robots by Learning Through Multimodal Feedback Channel which is organized by Cigdem Turan and Dorothea Koert together with Rudolf Lioutikov and Karl David Neergaard at the International Conference on Multimodal Interaction 2021 (October 18-22nd, 2021).
Workshop Introduction Paper

April 2021 We run through a first set of workshops with the associated industry partners Energy Robotics & Porsche Motorsport to define potential applications for interactive AI in industry. We look forward to the upcoming work on interesting topics together.

March 2021 The recent time has been filled with the elicitation of IKIDA use cases for research and experiments and the identification of potential value in industrial settings.

November 6, 2020 Inspiring official IKIDA kick-off – only virtually though, due to Corona.

October 2020 The BMBF-funded junior research group “IKIDA – Interaktive KI für Domänenexperten und Alltagsnutzer” has started on 01.10.2020. 6 young scientists will investigate interactive AI algorithms & cognitive models for human-AI interaction. Project duration 01.10.2020-30.09.2024 Learn more

  Name Working area(s) Contact
Dr.-Ing. Dorothea Koert
Head of IKIDA Team
Interactive Robot Skill Learning, Interactive Reinforcement Learning, Human Robot Interaction
+49 6151 16-20073
Vildan Salikutluk (Ph.D. Student)		Human Judgements and Problem Solving, Cognitive Modeling, Human-AI Interaction
+49 6151 16-24075
Lisa Scherf (Ph.D. Student)Human Robot Interaction, Interactive Robot Skill Learning
+49 6151 16-20073
Susanne Trick (Ph.D. Student)		Prediction of Human Behavior, Intention Recognition, Multimodal Human Robot Interaction, Classifier Fusion, Probabilistic Models
+49 6151 16-24171
João Carvalho (Ph.D. Student)		Reinforcement Learning, Robot Reinforcement Learning, Robotic Manipulation
+49-6151-16-25372
S2|02 E225
Dr.-Ing. Dirk Balfanz
Associated Researcher
Scientific Managing Director @ Centre for Cognitive Science / Focus in IKIDA: split of autonomy in interactions between humans and AI systems
+49 6151 16-23736
S1|15 245
Dr. Cigdem Turan
Postdoctoral Researcher
former team member
  • Salikutluk, V., Schöpper, J., Herbert, F., Scheuermann, K., Frodl, E., Balfanz, D., Jäkel, F. & Koert, D. (2024). An Evaluation of Situational Autonomy for Human-AI Collaboration in a Shared Workspace Setting. ACM Conference on Human Factors in Computing Systems (CHI) (accepted)
  • Scherf, L.; Gasche, L. A.; Chemangui, E.; Koert, D. (2024). Are you sure? – Multi-Modal Human Decision Uncertainty Detection in Human-Robot Interaction, 19th Annual ACM/IEEE International Conference on Human Robot Interaction (HRI) (accepted)
  • Carvalho, J.; Le, A.T.; Baierl, M.; Koert, D.; Peters, J. (2023). Motion Planning Diffusion: Learning and Planning of Robot Motions with Diffusion Models, IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS). PDF
  • Gassen, M.; Metzler, F.; Prescher, E.; Prasad, V.; Scherf, L. Kaiser, F.; Koert, D. (2023). I^3: Interactive Iterative Improvement for Few-Shot Action Segmentation, 32th IEEE International Conference on Robot & Human Interactive Communication (RO-MAN).
  • Salikutluk, V., Frodl, E., Herbert, F., Balfanz, D., Koert, D. (2023). Situational Adaptive Autonomy in Human-AI Cooperation. CHI’23 Workshop on “Automation XP23 – Intervening, Teaming, Delegating Creating Engaging Automation Experience”. PDF (opens in new tab)
  • Salikutluk, V., Koert, D., & Jäkel, F. (2023). Interacting with Large Language Models: A Case Study on AI-Aided Brainstorming for Guesstimation Problems. In HHAI 2023: Augmenting Human Intellect (pp. 153-167). IOS Press. PDF
  • Scherf, L.; Schmidt, A.; Pal, S.; Koert, D. (2023). Interactively Learning Behavior Trees from Imperfect Human Demonstrations (Extended Abstract), presented at Hybrid Human Artificial Intelligence HHAI'23.
  • Scherf, L.; Schmidt, A.; Pal, S.; Koert, D. (2023). Interactively Learning Behavior Trees from Imperfect Human Demonstrations, Frontiers in Robotics and AI.
  • Trick, S.; Lott, V.; Scherf, L.; Rothkopf, C. A.; Koert, D. (2023). What Can I Help You With: Towards Task-Independent Detection of Intentions for Interaction in a Human-Robot Environment. In 2023 32th IEEE International Conference on Robot & Human Interactive Communication (RO-MAN).
  • Carvalho, J.; Baierl, M; Urain, J; Peters, J. (2022). Conditioned Score-Based Models for Learning Collision-Free Trajectory Generation. NeurIPS 2022 Workshop on Score-Based Methods. PDF (opens in new tab)
  • Carvalho, J., Koert, D., Daniv, M., Peters, J. (2022). Adapting Object-Centric Probabilistic Movement Primitives with Residual Reinforcement Learning. Proceedings of the International Conference on Humanoid Robots (HUMANOIDS) 2022 PDF (opens in new tab) Video
  • Prasad, V., Koert, D., Stock-Homburg, R., Peters, J., Chalvatzaki, G. (2022). MILD: Multimodal Interactive Latent Dynamics for Learning Human-Robot Interaction. Proceedings of the International Conference on Humanoid Robots (HUMANOIDS) 2022 (accepted)
  • Rawal, N., Koert, D., Turan, C., Kersting, K., Peters, J., & Stock-Homburg, R. (2022). ExGenNet: Learning to Generate Robotic Facial Expression Using Facial Expression Recognition. Frontiers in Robotics and AI, 8. DOI PDF
  • Scherf, L.; Turan, C., Koert, D. (2022). Learning from Unreliable Human Action Advice in Interactive Reinforcement Learning. 2022 IEEE-RAS 21st International Conference on Humanoid Robots (Humanoids). PDF (opens in new tab)
  • Trick, S., Herbert, F., Rothkopf, C., & Koert, D. (2022). Interactive Reinforcement Learning With Bayesian Fusion of Multimodal Advice. IEEE Robotics and Automation Letters, 7(3), 7558-7565. DOI PDF (opens in new tab) Video
  • Trick, S., & Rothkopf, C. (2022). Bayesian Classifier Fusion with an Explicit Model of Correlation. Proceedings of The 25th International Conference on Artificial Intelligence and Statistics. DOI PDF (opens in new tab)
  • Knaust, M., & Koert, D. (2021). Guided Robot Skill Learning: A User-Study on Learning Probabilistic Movement Primitives with Non-Experts. In 2020 IEEE-RAS 20th International Conference on Humanoid Robots (Humanoids) (pp. 514-521). DOI PDF
  • Turan, C., Koert, D., Neergaard, K. D., & Lioutikov, R. (2021). Empowering Interactive Robots by Learning Through Multimodal Feedback Channel. In Proceedings of the 2021 International Conference on Multimodal Interaction. DOI PDF
  • M.Sc. Thesis, Jule Brendgen: The Relation between Social Interaction and Intrinsic Motivation in Reinforcement Learning, 2021
  • B.Sc. Thesis, Franziska Herbert: Using Multimodal Human Feedback for Reinforcement Learning, 2021
  • M.Sc. Thesis, Lisa Scherf: Learning to segment human sequential behavior to detect the intention for interaction, 2021
  • M.Sc. Thesis, Tümer Tosik: Reinforcement Learning and Implicit Feedback, 2021
  • B.Sc. Thesis, Simon Binz: The Role of Shielded Reinforcement Learning for Human-AI Interaction in a Collaborative Task Setting, 2022
  • M.Sc. Thesis, Julian Christopher Kerl: Analysing Implicit Facial Behaviour for Reinforcement Learning, 2022
  • B.Sc. Thesis, Marek Daniv: One Shot Graph Imitation Learning from Visual Data, 2022
  • B.Sc. Thesis, Erkam Ilhan: Analysis of Expert Knowledge as a Basis for AI-Aided Interactive Anomaly Detection, 2022 (in cooperation with Porsche Motorsport)
  • B. Sc. Thesis, Simon Jakoby: Interactive Anomaly Detection in Multivariate Timeseries Data, 2022
  • B.Sc. Thesis, Vilja Lott: Detecting an Intention for Interaction in a Human-Robot Environment, 2022
  • B.Sc. Thesis, Laura Sabioncello: The Role of Trust in AI in a Collaborative Game Setting, 2022
  • M.Sc. Thesis, Mattin Sayed: Solving Guesstimation Problems: Perceived Differences between Human and AI Support, 2022
  • B.Sc. Thesis, Nathalie Woortman: Comparing and Personalizing Human Following Behaviors for Mobile Ground Robots, 2022
  • M.Sc. Thesis, Chen Xue: Task Classification and Local Manipulation Controllers, 2022
  • B.Sc. Thesis, Isabelle Clev: An Exploration of Ethical Issues in Large Language Models, 2023
  • M.Sc. Thesis, Eric Frodl: Enhancing Human-AI Interaction with Probabilistic Behavior Prediction for Situational Autonomy Adaptation, 2023
  • B.Sc. Thesis, Jan Kegel: Solving Guesstimation Problems with a Visual Structuring Tool, 2023
  • M.Sc. Thesis, Jan Mackensen: A Human-Centered Approach for AI-Aided Anomaly Detection in Time Series, 2023
  • B.Sc. Thesis, Katrin Scheuermann: Human-AI Interaction: A Think Aloud Study on Cooperating with Autonomous Agents, 2023
  • B.Sc. Thesis, Janik Schöpper: Situational Autonomy in Human-AI Interaction, 2023
  • B.Sc. Thesis, Sophie Schumbert: The Presentation of Uncertainty in Human-AI Interaction, 2023

We are always searching for good students and can offer interesting HiWi positions for students of the TU Darmstadt.
Open Topics:

  • Theory of Mind Models and Dynamic Adaptive Autonomy for HRI in Shared Workspaces (BT/MT/HIWI)
  • Interactive Anomaly Detection in Multivariate Timeseries Data (BT/MT)
  • Interactive Robot Skill Learning (BT/MT)
  • Intrinsic Motivation and Interactive Reinforcement Learning (BT/MT)
  • Teaching a Robot to play Ubongo 3D (BT/MT)

If you are interested in a bachelor or master thesis or a HiWi job with us, please contact
Beside these topics we are also open to suggestions by students in terms of Master's and Bachelor's Theses as long as they touch topics relevant to IKIDA.

Project Details

Project: IKIDA – Interaktive KI für Domänenexperten und Alltagsnutzer
Project partners: Technical University of Darmstadt (TU Darmstadt)
Project duration: October 2020 – September 2024
Project funding: 1.95 Mio EUR
Funded by: German Federal Ministry of Education and Research (BMBF)
Grant no.: 01IS20045