Dr. David Obremski

Dr. David Obremski

Senior Researcher

Tel.: +49 931 31 88144

Psychologie Intelligenter Interaktiver Systeme
Julius-Maximilians-Universität Würzburg
Institut für Mensch-Computer-Medien
Universität Würzburg
Emil-Fischer-Straße 50
97074 Würzburg
Room 01.002

Mini-CV

David completed his Bachelor's degree in Human-Computer Systems and his Master's degree in Human-Computer Interaction in Würzburg. At the University of Music in Würzburg, David studied jazz piano in his Bachelor's degree. He wrote his PhD thesis in the field of mixed-cultural virtual agents and their possible applications for the reduction of implicit cultural bias at the work group for media informatics, supervised by Prof. Birgit Lugrin. Since April 2023, David has been working in the Psychology of Intelligent Interactive Systems group. His research focuses on the perception of (embodied) artificial intelligence with an emphasis on user trust, effectiveness of human-ai collaborative work settings and psychotherapy and AI.

Research Interests

Artificial Intelligence (AI)
Embodied AI
Extended Reality
AI in Psychotherapy
AI in Work Environments
Trust in AI

Publications

2026[ to top ]

Kern, Florian, Lukas Polifke, Paula Friedrich, Marc Erich Latoschik, Carolin Wienrich, and David Obremski. “CECA - A Configurable Framework for Embodied Conversational AI Agents in Extended Reality”. In 2026 IEEE Conference on Virtual Reality and 3D User Interfaces Abstracts and Workshops (VRW).
- [ Abstract ]
- [ BibTeX ]
We present CECA, a configurable framework for embodied conversational AI agents in Unity-based extended reality (XR) applications. CECA employs a client–server architecture to decouple agent logic from game engine–based embodiment. Built on LiveKit Agents, our approach integrates speech-to-text (STT), large language models (LLMs), and text-to-speech (TTS) into a unified, streaming voice-to-voice pipeline configured via metadata rather than code changes. We outline how this architecture flexibly integrates local and cloud AI providers while mitigating limited provider SDK support in Unity. Finally, we highlight opportunities for future work, including multi-agent scenarios, higher-level templates for XR research, and systematic user studies.

@inproceedings{kern2026configurable, abstract = {We present CECA, a configurable framework for embodied conversational AI agents in Unity-based extended reality (XR) applications. CECA employs a client–server architecture to decouple agent logic from game engine–based embodiment. Built on LiveKit Agents, our approach integrates speech-to-text (STT), large language models (LLMs), and text-to-speech (TTS) into a unified, streaming voice-to-voice pipeline configured via metadata rather than code changes. We outline how this architecture flexibly integrates local and cloud AI providers while mitigating limited provider SDK support in Unity. Finally, we highlight opportunities for future work, including multi-agent scenarios, higher-level templates for XR research, and systematic user studies.}, author = {Kern, Florian and Polifke, Lukas and Friedrich, Paula and Latoschik, Marc Erich and Wienrich, Carolin and Obremski, David}, booktitle = {2026 IEEE Conference on Virtual Reality and 3D User Interfaces Abstracts and Workshops (VRW)}, keywords = {obremski}, note = {To be published}, title = {CECA - A Configurable Framework for Embodied Conversational AI Agents in Extended Reality}, year = 2026 }
Obremski, David, Paula Friedrich, and Carolin Wienrich. “To Be Healed or Hacked? - User‑Centered Ethical Design for Embodied AI in Mental Health Care”. IEEE Transactions on Visualization and Computer Graphics, 2026.
- [ Abstract ]
- [ BibTeX ]
The global prevalence of mental health disorders has created a substantial treatment gap. To support clinicians and increase access to care, researchers in the field of Artificial Intelligence (AI) and Virtual Reality (VR) have investigated technology-mediated psychotherapy for years. However, research about stakeholders' concerns and their readiness to use AI in psychotherapy remains scarce. This study focuses on a user-centered approach to accommodate patients' concerns and, based on the results, implement measures to foster self-disclosure and trust towards an embodied AI therapist in VR. First, we conducted an online study with mental health patients ($N = 152$), which identified data autonomy and transparency as their primary ethical concerns. In a subsequent in-person VR study ($N = 90$) we compared effects of increased data autonomy and transparency on self-disclosure and trust towards an embodied AI therapist. Results indicated that higher data autonomy led to greater self-disclosure, while transparency had no significant effect. Manipulating data autonomy and transparency did not affect perceived trust, though exploratory calculations revealed that women reported significantly higher trust levels than men. These findings illuminate patients' priorities and provide implications for technical designs for AI-driven mental health care.

@article{obremski2026healed, abstract = {The global prevalence of mental health disorders has created a substantial treatment gap. To support clinicians and increase access to care, researchers in the field of Artificial Intelligence (AI) and Virtual Reality (VR) have investigated technology-mediated psychotherapy for years. However, research about stakeholders' concerns and their readiness to use AI in psychotherapy remains scarce. This study focuses on a user-centered approach to accommodate patients' concerns and, based on the results, implement measures to foster self-disclosure and trust towards an embodied AI therapist in VR. First, we conducted an online study with mental health patients ($N = 152$), which identified data autonomy and transparency as their primary ethical concerns. In a subsequent in-person VR study ($N = 90$) we compared effects of increased data autonomy and transparency on self-disclosure and trust towards an embodied AI therapist. Results indicated that higher data autonomy led to greater self-disclosure, while transparency had no significant effect. Manipulating data autonomy and transparency did not affect perceived trust, though exploratory calculations revealed that women reported significantly higher trust levels than men. These findings illuminate patients' priorities and provide implications for technical designs for AI-driven mental health care.}, author = {Obremski, David and Friedrich, Paula and Wienrich, Carolin}, journal = {IEEE Transactions on Visualization and Computer Graphics}, keywords = {ieeevr26}, note = {To be published}, title = {To be Healed or Hacked? - User‑Centered Ethical Design for Embodied AI in Mental Health Care}, year = 2026 }

2025[ to top ]

Krop, Philipp, David Obremski, Astrid Carolus, Marc Erich Latoschik, and Carolin Wienrich. “My Co-Worker ChatGPT: Development of an XR Application for Embodied Artificial Intelligence in Work Environments”. In 2025 IEEE Conference on Virtual Reality and 3D User Interfaces Abstracts and Workshops (IEEE VRW). IEEE Computer Science, 2025. https://doi.org/10.1109/VRW66409.2025.00468.
- [ Abstract ]
- [ BibTeX ]
- [ URL ]
- [ DOI ]
With recent developments in spatial computing, work contexts might shift to augmented reality. Embodied AI - virtual conversa- tional agents backed by AI systems - have the potential to enhance these contexts and open up more communication channels than just text. To support knowledge transfer from virtual agent research to the general populace, we developed My CoWorker ChatGPT - an interactive demo where employees can try out various embodied AIs in a virtual office or their own using augmented reality. We use state-of-the-art speech synthesis and body-scanning technology to create believable and trustworthy AI assistants. The demo was shown at multiple events throughout Germany, where it was well received and sparked fruitful conversations about the possibilities of embodied AI in work contexts.

@inproceedings{krop2025coworker, abstract = {With recent developments in spatial computing, work contexts might shift to augmented reality. Embodied AI - virtual conversa- tional agents backed by AI systems - have the potential to enhance these contexts and open up more communication channels than just text. To support knowledge transfer from virtual agent research to the general populace, we developed My CoWorker ChatGPT - an interactive demo where employees can try out various embodied AIs in a virtual office or their own using augmented reality. We use state-of-the-art speech synthesis and body-scanning technology to create believable and trustworthy AI assistants. The demo was shown at multiple events throughout Germany, where it was well received and sparked fruitful conversations about the possibilities of embodied AI in work contexts.}, author = {Krop, Philipp and Obremski, David and Carolus, Astrid and Latoschik, Marc Erich and Wienrich, Carolin}, booktitle = {2025 IEEE Conference on Virtual Reality and 3D User Interfaces Abstracts and Workshops (IEEE VRW)}, keywords = {aiatwork}, month = {03}, publisher = {IEEE Computer Science}, title = {My Co-worker ChatGPT: Development of an XR Application for Embodied Artificial Intelligence in Work Environments}, year = 2025 }
Wienrich, Carolin, and David Obremski. “Immersion, Attention, and Collaboration in Spatial Computing: A Study on Work Performance With Apple Vision Pro”. IEEE Transactions on Visualization and Computer Graphics 31, no. 5 (2025): 2995-3002. https://doi.org/10.1109/TVCG.2025.3549145.
- [ Abstract ]
- [ BibTeX ]
- [ DOI ]
Spatial computing is set to change the way we work. It will enable both focused work through a higher degree of immersion and collaborative work through enhanced integration of shared interaction spaces or interaction partners. With the Apple Vision Pro, the level of immersion can be adjusted seamlessly. So far, there have been no systematic studies on how this adjustability affects work performance when working alone or together. The present empirical study fills this research gap by varying the level of immersion across three stages (high, medium, low) while solving various tasks with the Apple Vision Pro. The results show that selective attention improves significantly with increasing immersion levels. In contrast, social presence decreases with increasing immersion. In general, participants performed better in the individual task than in the collaborative task. However, the degree of immersion did not influence the collaborative performance. In addition, we could not determine any adverse effects on depth perception or user experience after use. The present study provides initial contributions to the future of spatial computing in professional settings and highlights the importance of balancing immersion and social interaction in a world where digital and physical spaces seamlessly coexist

@article{10917012, abstract = {Spatial computing is set to change the way we work. It will enable both focused work through a higher degree of immersion and collaborative work through enhanced integration of shared interaction spaces or interaction partners. With the Apple Vision Pro, the level of immersion can be adjusted seamlessly. So far, there have been no systematic studies on how this adjustability affects work performance when working alone or together. The present empirical study fills this research gap by varying the level of immersion across three stages (high, medium, low) while solving various tasks with the Apple Vision Pro. The results show that selective attention improves significantly with increasing immersion levels. In contrast, social presence decreases with increasing immersion. In general, participants performed better in the individual task than in the collaborative task. However, the degree of immersion did not influence the collaborative performance. In addition, we could not determine any adverse effects on depth perception or user experience after use. The present study provides initial contributions to the future of spatial computing in professional settings and highlights the importance of balancing immersion and social interaction in a world where digital and physical spaces seamlessly coexist}, author = {Wienrich, Carolin and Obremski, David}, journal = {IEEE Transactions on Visualization and Computer Graphics}, keywords = {ieeevr25}, number = 5, pages = {2995-3002}, title = {Immersion, Attention, and Collaboration in Spatial Computing: a Study on Work Performance with Apple Vision Pro}, volume = 31, year = 2025 }
Obremski, David, Paula Friedrich, and Carolin Wienrich. “Please Let Me Think: The Influence of Conversational Fillers on Transparency and Perception of Waiting Time When Interacting With a Conversational AI in Virtual Reality”. In Proceedings of the 27th International Conference on Multimodal Interaction, 496-505. http://dblp.uni-trier.de/db/conf/icmi/icmi2025.html#ObremskiFW25.
- [ BibTeX ]
- [ URL ]
@inproceedings{obremski2025please, author = {Obremski, David and Friedrich, Paula and Wienrich, Carolin}, booktitle = {Proceedings of the 27th International Conference on Multimodal Interaction}, keywords = {myown}, pages = {496–505}, title = {Please Let Me Think: The Influence of Conversational Fillers on Transparency and Perception of Waiting Time when Interacting with a Conversational AI in Virtual Reality}, year = 2025 }

2024[ to top ]

Obremski, David, and Carolin Wienrich. “Autonomous VR Exposure Therapy for Anxiety Disorders Using Conversational AI—Ethical and Technical Implications and Preliminary Results”. Scientific Research Publishing 12, no. 4 (April 2024). https://doi.org/10.4236/jcc.2024.124010.
- [ Abstract ]
- [ BibTeX ]
- [ URL ]
- [ DOI ]
The need for psychotherapy is very high and the lack of care causes a lot of suffering and high costs. This paper presents an interdisciplinary approach to creating an AI-guided exposure therapy for fear of heights in virtual reality (VR). First, ethical principles for the use of conversational AI in psychotherapy were translated into technical requirements and made measurable. Based on this, an autonomous virtual reality exposure therapy was iteratively developed with a therapist. The feasibility and implementation of the ethical principles were tested with a patient. The patient was very satisfied with the VR setup. The AI therapist was also rated positively, although there is still room for improvement regarding conversational skills. Overall, the paper shows how AI can contribute responsibly to improving the psycho-therapeutic supply. It also provides guidelines that make ethical principles tangible and measurable for developers.

@article{obremski2024autonomous, abstract = {The need for psychotherapy is very high and the lack of care causes a lot of suffering and high costs. This paper presents an interdisciplinary approach to creating an AI-guided exposure therapy for fear of heights in virtual reality (VR). First, ethical principles for the use of conversational AI in psychotherapy were translated into technical requirements and made measurable. Based on this, an autonomous virtual reality exposure therapy was iteratively developed with a therapist. The feasibility and implementation of the ethical principles were tested with a patient. The patient was very satisfied with the VR setup. The AI therapist was also rated positively, although there is still room for improvement regarding conversational skills. Overall, the paper shows how AI can contribute responsibly to improving the psycho-therapeutic supply. It also provides guidelines that make ethical principles tangible and measurable for developers.}, author = {Obremski, David and Wienrich, Carolin}, journal = {Scientific Research Publishing}, keywords = {obremski}, month = {04}, number = 4, title = {Autonomous VR Exposure Therapy for Anxiety Disorders Using Conversational AI—Ethical and Technical Implications and Preliminary Results}, volume = 12, year = 2024 }
Wienrich, Carolin, Marc Erich Latoschik, and David Obremski. “Gender Differences and Social Design in Human-AI Collaboration: Insights from Virtual Cobot Interactions Under Varying Task Loads”. In Extended Abstracts of the 2024 CHI Conference on Human Factors in Computing Systems. CHI EA ’24. <conf-loc> <city>Honolulu</city> <state>HI</state> <country>USA</country> </conf-loc>: Association for Computing Machinery, 2024. https://doi.org/10.1145/3613905.3650827.
- [ Abstract ]
- [ BibTeX ]
- [ URL ]
- [ DOI ]
This work explores the effects of users’ gender and social design features of AI under different task load conditions on human-like attributions, social impact, work performance and perceived workload, user experience, and various other measures in Human-AI Interaction (HAII). Users had to execute sorting and dispatch tasks in collaboration with a virtual cobot. The degree of social gestalt of the cobot was varied by the ability to make small talk (i.e., talkative vs. non-talkative cobot), and the task load was increased by adding a secondary task (i.e., high vs. low task load condition). Overall, the talkative cobot led to a more positive perception of the cobot and increased social qualities like sense of meaning and team membership compared to the non-talkative cobot. The following gender effect was particularly interesting. The talkative cobot had a buffering effect for women and a distraction conflict effect for men in high task load conditions. When interacting with the talkative robot, women find the high task condition less stressful. In contrast thereto, the talkative cobot was distracting for men in the high task load condition. Our results highlight that social design choices and interindividual differences influence a successful collaboration between humans and AI. The work also shows the added value of systematic XR-simulations for the investigation and design of human-centered HAIIs (eXtended AI approach).

@inproceedings{10.1145/3613905.3650827, abstract = {This work explores the effects of users’ gender and social design features of AI under different task load conditions on human-like attributions, social impact, work performance and perceived workload, user experience, and various other measures in Human-AI Interaction (HAII). Users had to execute sorting and dispatch tasks in collaboration with a virtual cobot. The degree of social gestalt of the cobot was varied by the ability to make small talk (i.e., talkative vs. non-talkative cobot), and the task load was increased by adding a secondary task (i.e., high vs. low task load condition). Overall, the talkative cobot led to a more positive perception of the cobot and increased social qualities like sense of meaning and team membership compared to the non-talkative cobot. The following gender effect was particularly interesting. The talkative cobot had a buffering effect for women and a distraction conflict effect for men in high task load conditions. When interacting with the talkative robot, women find the high task condition less stressful. In contrast thereto, the talkative cobot was distracting for men in the high task load condition. Our results highlight that social design choices and interindividual differences influence a successful collaboration between humans and AI. The work also shows the added value of systematic XR-simulations for the investigation and design of human-centered HAIIs (eXtended AI approach).}, address = {New York, NY, USA}, author = {Wienrich, Carolin and Latoschik, Marc Erich and Obremski, David}, booktitle = {Extended Abstracts of the 2024 CHI Conference on Human Factors in Computing Systems}, keywords = {obremski}, publisher = {Association for Computing Machinery}, series = {CHI EA '24}, title = {Gender Differences and Social Design in Human-AI Collaboration: Insights from Virtual Cobot Interactions Under Varying Task Loads}, year = 2024 }
Wienrich, Carolin, Stephanie Vogt, Nina Döllinger, and David Obremski. “Promoting Eco-Friendly Behavior through Virtual Reality - Implementation and Evaluation of Immersive Feedback Conditions of a Virtual CO2 Calculator”. In 2024 CHI Conference on Human Factors in Computing Systems. https://downloads.hci.informatik.uni-wuerzburg.de/2024_chi_co2_calculator.pdf.
- [ Abstract ]
- [ BibTeX ]
- [ URL ]
Climate change is one of the most pressing global challenges in the 21st century. Urgent actions favoring the environment's well-being are essential to mitigate its potentially irreversible consequences. However, the delayed and often distant nature of the effects of sustainable behavior makes it challenging for individuals to connect with the issue personally. Immersive media are an opportunity to introduce innovative feedback mechanisms to highlight the urgency of behavior effects. We introduce a VR carbon calculator that visualizes users' annual carbon footprint as CO2-filled balloons over multiple periods. In a 2x2 design, participants calculated and visualized their carbon footprint numerically or as balloons over one or three years. We found no effect of our visualization but a significant impact of the visualized period on participants' environmental self-efficacy. These findings emphasize the importance of target-oriented design in VR behavior interventions.

@inproceedings{wienrich2024promoting, abstract = {Climate change is one of the most pressing global challenges in the 21st century. Urgent actions favoring the environment's well-being are essential to mitigate its potentially irreversible consequences. However, the delayed and often distant nature of the effects of sustainable behavior makes it challenging for individuals to connect with the issue personally. Immersive media are an opportunity to introduce innovative feedback mechanisms to highlight the urgency of behavior effects. We introduce a VR carbon calculator that visualizes users' annual carbon footprint as CO2-filled balloons over multiple periods. In a 2x2 design, participants calculated and visualized their carbon footprint numerically or as balloons over one or three years. We found no effect of our visualization but a significant impact of the visualized period on participants' environmental self-efficacy. These findings emphasize the importance of target-oriented design in VR behavior interventions.}, author = {Wienrich, Carolin and Vogt, Stephanie and Döllinger, Nina and Obremski, David}, booktitle = {2024 CHI Conference on Human Factors in Computing Systems}, keywords = {c.wienrich}, title = {Promoting Eco-Friendly Behavior through Virtual Reality - Implementation and Evaluation of Immersive Feedback Conditions of a Virtual CO2 Calculator}, year = 2024 }

2022[ to top ]

Obremski, David, Paula Friedrich, Nora Haak, Philipp Schaper, and Birgit Lugrin. “The Impact of Mixed-Cultural Speech on the Stereotypical Perception of a Virtual Robot”. Frontiers in Robotics and AI 9 (November 2022). https://doi.org/10.3389/frobt.2022.983955.
- [ BibTeX ]
- [ URL ]
- [ DOI ]
@article{Obremski_2022, author = {Obremski, David and Friedrich, Paula and Haak, Nora and Schaper, Philipp and Lugrin, Birgit}, journal = {Frontiers in Robotics and AI}, keywords = {accent}, month = 11, publisher = {Frontiers Media SA}, title = {The impact of mixed-cultural speech on the stereotypical perception of a virtual robot}, volume = 9, year = 2022 }
Wienrich, Carolin, David Obremski, and Johann Habakuk Israel. “Repeated Experience or a Virtual Nose to Reduce Simulator Sickness? Investigating Prediction of the Sensorial Conflict Theory and the Rest-Frame Hypothesis in Two Virtual Games”. Entertainment Computing 43 (August 2022): 100514. https://doi.org/10.1016/j.entcom.2022.100514.
- [ BibTeX ]
- [ URL ]
- [ DOI ]
@article{Wienrich_2022, author = {Wienrich, Carolin and Obremski, David and Israel, Johann Habakuk}, journal = {Entertainment Computing}, keywords = {MI}, month = {08}, pages = 100514, publisher = {Elsevier {BV}}, title = {Repeated experience or a virtual nose to reduce simulator sickness? Investigating prediction of the sensorial conflict theory and the rest-frame hypothesis in two virtual games}, volume = 43, year = 2022 }
Obremski, David, Helena Babette Hering, Paula Friedrich, and Birgit Lugrin. “Mixed-Cultural Speech for Intelligent Virtual Agents - the Impact of Different Non-Native Accents Using Natural or Synthetic Speech in the English Language”. In HAI ’22: International Conference on Human-Agent Interaction, 67-75. ACM, 2022. https://doi.org/10.1145/3527188.3561921.
- [ BibTeX ]
- [ URL ]
- [ DOI ]
@inproceedings{Obremski_HAI_Paper_2022, author = {Obremski, David and Hering, Helena Babette and Friedrich, Paula and Lugrin, Birgit}, booktitle = {HAI '22: International Conference on Human-Agent Interaction}, keywords = {accent}, month = 12, pages = {67-75}, publisher = {ACM}, title = {Mixed-Cultural Speech for Intelligent Virtual Agents - the Impact of Different Non-Native Accents Using Natural or Synthetic Speech in the English Language}, year = 2022 }
Obremski, David, Helena Babette Hering, Paula Friedrich, and Birgit Lugrin. “Exploratory Study on the Perception of Intelligent Virtual Agents With Non-Native Accents Using Synthetic and Natural Speech in German”. In INTERNATIONAL CONFERENCE ON MULTIMODAL INTERACTION. ACM, 2022. https://doi.org/10.1145/3536221.3556608.
- [ BibTeX ]
- [ URL ]
- [ DOI ]
@inproceedings{Obremski_2022, author = {Obremski, David and Hering, Helena Babette and Friedrich, Paula and Lugrin, Birgit}, booktitle = {INTERNATIONAL CONFERENCE ON MULTIMODAL INTERACTION}, keywords = {accent}, month = 11, publisher = {ACM}, title = {Exploratory Study on the Perception of Intelligent Virtual Agents With Non-Native Accents Using Synthetic and Natural Speech in German}, year = 2022 }
Obremski, David, and Birgit Lugrin. “Mixed-Cultural Speech for Mixed-Cultural Users - Natural Vs. Synthetic Speech for Virtual Agents”. In HAI ’22: International Conference on Human-Agent Interaction, 290-92. ACM, 2022. https://doi.org/10.1145/3527188.3563910.
- [ BibTeX ]
- [ URL ]
- [ DOI ]
@inproceedings{Obremski_HAI_Poster_2022, author = {Obremski, David and Lugrin, Birgit}, booktitle = {HAI '22: International Conference on Human-Agent Interaction}, keywords = {accent}, month = 12, pages = {290-292}, publisher = {ACM}, title = {Mixed-Cultural Speech for Mixed-Cultural Users - Natural vs. Synthetic Speech for Virtual Agents}, year = 2022 }

2021[ to top ]

Obremski, David, Alicia L. Schäfer, Benjamin P. Lange, Birgit Lugrin, Elisabeth Ganal, Laura Witt, Tania R. Nuñez, Sascha Schwarz, and Frank Schwab. “Put That Away and Talk to Me - the Effects of Smartphone Induced Ostracism While Interacting With an Intelligent Virtual Agent”. In Proceedings of the 9th International Conference on Human-Agent Interaction 2021. ACM, 2021. https://doi.org/10.1145/3472307.3484669.
- [ BibTeX ]
- [ DOI ]
@inproceedings{Obremski_2021, author = {Obremski, David and Schäfer, Alicia L. and Lange, Benjamin P. and Lugrin, Birgit and Ganal, Elisabeth and Witt, Laura and Nuñez, Tania R. and Schwarz, Sascha and Schwab, Frank}, booktitle = {Proceedings of the 9th International Conference on Human-Agent Interaction 2021}, keywords = {mi}, publisher = {ACM}, title = {Put that Away and Talk to Me - the Effects of Smartphone induced Ostracism while Interacting with an Intelligent Virtual Agent}, year = 2021 }
Obremski, David, Jean-Luc Lugrin, Philipp Schaper, and Birgit Lugrin. “Non-Native Speaker Perception of Intelligent Virtual Agents in Two Languages: The Impact of Amount and Type of Grammatical Mistakes”. Journal on Multimodal User Interfaces, 2021. https://doi.org/10.1007/s12193-021-00369-9.
- [ Abstract ]
- [ BibTeX ]
- [ URL ]
- [ DOI ]
Having a mixed-cultural membership becomes increasingly common in our modern society. It is thus beneficial in several ways to create Intelligent Virtual Agents (IVAs) that reflect a mixed-cultural background as well, e.g., for educational settings. For research with such IVAs, it is essential that they are classified as non-native by members of a target culture. In this paper, we focus on variations of IVAs' speech to create the impression of non-native speakers that are identified as such by speakers of two different mother tongues. In particular, we investigate grammatical mistakes and identify thresholds beyond which the agents is clearly categorised as a non-native speaker. Therefore, we conducted two experiments: one for native speakers of German, and one for native speakers of English. Results of the German study indicate that beyond 10{\%} of word order mistakes and 25{\%} of infinitive mistakes German-speaking IVAs are perceived as non-native speakers. Results of the English study indicate that beyond 50{\%} of omission mistakes and 50{\%} of infinitive mistakes English-speaking IVAs are perceived as non-native speakers. We believe these thresholds constitute helpful guidelines for computational approaches of non-native speaker generation, simplifying research with IVAs in mixed-cultural settings.

@article{Obremski2021, abstract = {Having a mixed-cultural membership becomes increasingly common in our modern society. It is thus beneficial in several ways to create Intelligent Virtual Agents (IVAs) that reflect a mixed-cultural background as well, e.g., for educational settings. For research with such IVAs, it is essential that they are classified as non-native by members of a target culture. In this paper, we focus on variations of IVAs' speech to create the impression of non-native speakers that are identified as such by speakers of two different mother tongues. In particular, we investigate grammatical mistakes and identify thresholds beyond which the agents is clearly categorised as a non-native speaker. Therefore, we conducted two experiments: one for native speakers of German, and one for native speakers of English. Results of the German study indicate that beyond 10{\%} of word order mistakes and 25{\%} of infinitive mistakes German-speaking IVAs are perceived as non-native speakers. Results of the English study indicate that beyond 50{\%} of omission mistakes and 50{\%} of infinitive mistakes English-speaking IVAs are perceived as non-native speakers. We believe these thresholds constitute helpful guidelines for computational approaches of non-native speaker generation, simplifying research with IVAs in mixed-cultural settings.}, author = {Obremski, David and Lugrin, Jean-Luc and Schaper, Philipp and Lugrin, Birgit}, journal = {Journal on Multimodal User Interfaces}, keywords = {mi}, title = {Non-native speaker perception of Intelligent Virtual Agents in two languages: the impact of amount and type of grammatical mistakes}, year = 2021 }

2020[ to top ]

Obremski, David, Carolin Wienrich, and Astrid Carolus. “What the user’s Voice Tells Us about UX - Analysing Parameters of the Voice As Indicators of the User Experience of the Usage of Intelligent Voice Assistants”. In. Gesellschaft für Informatik e.V, 2020. https://doi.org/10.18420/MUC2020-WS105-377.
- [ BibTeX ]
- [ URL ]
- [ DOI ]
@inproceedings{https://doi.org/10.18420/muc2020-ws105-377, author = {Obremski, David and Wienrich, Carolin and Carolus, Astrid}, keywords = {obremski}, publisher = {Gesellschaft für Informatik e.V.}, title = {What the user’s voice tells us about UX - Analysing parameters of the voice as indicators of the User Experience of the usage of intelligent voice assistants}, year = 2020 }

2019[ to top ]

Obremski, David, Jean-Luc Lugrin, Philipp Schaper, and Birgit Lugrin. “Non-Native Speaker Generation and Perception for Mixed-Cultural Settings”. ACM International Conference on Intelligent Virtual Agents 2019, 2019. https://doi.org/https://doi.org/10.1145/3308532.3329427.
- [ Abstract ]
- [ BibTeX ]
- [ DOI ]
This paper presents an experiment evaluating the effects of virtual agents' language proficiency on whether they are perceived as a native speakers or not. Our first results indicate that beyond 10% word order mistakes and 25% infinitive mistakes, virtual agents are perceived as non-native speakers, even though their appearance and non-verbal behaviour were not altered. We believe these thresholds constitute interesting guidelines possibly simplifying the design of non-native speaker simulations.

@article{obremski2019nonnative, abstract = {This paper presents an experiment evaluating the effects of virtual agents' language proficiency on whether they are perceived as a native speakers or not. Our first results indicate that beyond 10% word order mistakes and 25% infinitive mistakes, virtual agents are perceived as non-native speakers, even though their appearance and non-verbal behaviour were not altered. We believe these thresholds constitute interesting guidelines possibly simplifying the design of non-native speaker simulations.}, author = {Obremski, David and Lugrin, Jean-Luc and Schaper, Philipp and Lugrin, Birgit}, journal = {ACM International Conference on Intelligent Virtual Agents 2019}, keywords = {mi}, publisher = {ACM}, title = {Non-Native Speaker Generation and Perception for Mixed-Cultural Settings}, year = 2019 }

2016[ to top ]

Lugrin, Jean-Luc, David Obremski, Daniel Roth, and Marc Erich Latoschik. “Audio Feedback and Illusion of Virtual Body Ownership in Mixed Reality”. In Proceedings of the 22Nd ACM Conference on Virtual Reality Software and Technology, 309-10. VRST \u002716. Munich, Germany: ACM, 2016. https://doi.org/10.1145/2993369.2996319.
- [ BibTeX ]
- [ URL ]
- [ DOI ]
@inproceedings{Lugrin:2016:AFI:2993369.2996319, address = {New York, NY, USA}, author = {Lugrin, Jean-Luc and Obremski, David and Roth, Daniel and Latoschik, Marc Erich}, booktitle = {Proceedings of the 22Nd ACM Conference on Virtual Reality Software and Technology}, keywords = {ivbo}, pages = {309-310}, publisher = {ACM}, series = {VRST \u002716}, title = {Audio Feedback and Illusion of Virtual Body Ownership in Mixed Reality}, year = 2016 }

Senior Researcher

Mini-CV

Research Interests

Bildnachweise