Workshop “Generative AI meets Eye Tracking (GenEAI)”
ACM Symposium on Eye Tracking Research and Applications (ACM ETRA 2026)
June 1-4, 2026, Marrakesh, Morocco
Generative artificial intelligence has come into focus with applications in content generation, design, and predictive modeling. At the same time, eye-tracking technologies have provided unparalleled insights into human attention, perception, and behavior, enabling various applications in neuroscience, human-computer interaction (HCI), and marketing. Bringing these two powerful technologies together will open up new possibilities for understanding and improving human interaction with AI systems.
This workshop aims to explore the synergies between these two cutting-edge fields and provide a platform for researchers, practitioners, and developers to share new approaches, tools, and methods at the intersection of these fields. The workshop will also address possible ethical considerations and challenges that arise from this integration, including privacy concerns and the implications of using data on human gaze behavior as a source of information for generative algorithms.
Submission
We invite authors to submit original work following the ETRA SHORT PAPER format (max. 8 pages, plus any number of additional pages for references, with a 150-word abstract). Manuscripts should emphasize the intersection of generative AI and eye tracking, including any of the topics listed above. Please ensure that your submission complies with the Author Guidelines for SIG-sponsored events (sigconf format). The submission and review process will be handled through the Precision Conference System (PCS). All accepted papers will be published by ACM as part of the Workshop Proceedings of ETRA 2026.
Important Dates
| Event | Date |
|---|---|
| Submission Deadline | March 1, 2026 |
| Notification | March 20, 2026 |
| Camera Ready Deadline | March 30, 2026 |
Topics of Interest
- Generative AI models informed by eye-tracking data for improved content generation and personalization.
- Foundation models and their application to analyze eye-tracking data in new and efficient ways.
- Real-time adaptive interfaces that combine generative AI and eye tracking for dynamic user experiences in AR/VR, gaming, and education.
- Personalization of user interfaces and experiences based on gaze data.
- Behavioral and cognitive insights gained by analyzing gaze data with generative AI, with applications in HCI, marketing, and psychology.
- Visualization techniques for eye movement data, including spatio-temporal analysis and visual exploration of gaze patterns.
- Ethical and privacy considerations in using eye-tracking data to inform generative AI models.
- Applications of eye-tracking in challenging environments, including mobile devices, large displays, and mixed/virtual reality systems.
- Integration of eye-tracking into LLMs/VLMs for more intuitive and responsive human-AI interaction systems.
Preliminary Program
Workshop “Generative AI meets Eye Tracking (GenEAI)” will take place during ACM ETRA 2026 (June 1-4, 2026, Marrakesh, Morocco).
- Keynote 1: Yue Jiang (University of Utah) — The Synthetic User: Simulating Viewing Behaviors through Multimodal AI.
- Paper presentations Session 1
- Break
- Keynote 2: Zoya Bylinskii (Adobe Firefly) — Perceptual Insights for GenAI Evaluation
- Paper presentations Session 2
- Lunch/Break
- Paper presentations Session 3
- Keynote 4: Vaibhav Unhelkar (Rice University) — Training Humans to Work with Robots
Keynote Speakers
Yue Jiang, University of Utah
The Synthetic User: Simulating Viewing Behaviors through Multimodal AI.
Understanding how users visually process information is critical for effective interface design, yet gathering eye-tracking data remains resource-intensive. The rise of multimodal generative AI presents a compelling alternative: the creation of “synthetic users” that can simulate human viewing behaviors with high fidelity. This talk shows the capabilities of state-of-the-art generative models to predict visual attention on Graphical User Interfaces (GUIs). Unlike traditional saliency maps, these multimodal agents account for specific user goals and GUI layouts, generating sequential fixation patterns that mirror human viewing strategies. The talk will cover the technical nuances of training these models on large-scale eye-tracking datasets, the evaluation metrics required to assess generative scanpaths, and the current limitations regarding individual differences. Yue Jiang is an Assistant Professor at the University of Utah, where she leads research at the intersection of Human-Computer Interaction and Computer Vision. Her work focuses on Computational User Interface Understanding, with specific expertise in applying multimodal generative AI to model human viewing behavior and generate adaptive, accessible UIs. Yue’s research has been published in top venues including CHI, UIST, and CVPR. She initiated and served as the lead organizer of five Computational UI workshops at CHI (2022–2026) and co-organized the Mobile Eye-based Interaction Workshop at ETRA. She was the Accessibility Chair for CHI 2023 and CHI 2024 and will be the Workshop Chair for IUI 2027. Her contributions have been recognized with the Meta Research PhD Fellowship, the Google Europe Scholarship, and the Nokia Scholarship. She was also selected as a Heidelberg Laureate Forum Young Researcher and a UCSD Data Science Rising Star. Previously, Yue has worked or interned at Intel, Adobe Research, Apple, the Max Planck Institute, and Carnegie Mellon University.
The Synthetic User: Simulating Viewing Behaviors through Multimodal AI.
Understanding how users visually process information is critical for effective interface design, yet gathering eye-tracking data remains resource-intensive. The rise of multimodal generative AI presents a compelling alternative: the creation of “synthetic users” that can simulate human viewing behaviors with high fidelity. This talk shows the capabilities of state-of-the-art generative models to predict visual attention on Graphical User Interfaces (GUIs). Unlike traditional saliency maps, these multimodal agents account for specific user goals and GUI layouts, generating sequential fixation patterns that mirror human viewing strategies. The talk will cover the technical nuances of training these models on large-scale eye-tracking datasets, the evaluation metrics required to assess generative scanpaths, and the current limitations regarding individual differences. Yue Jiang is an Assistant Professor at the University of Utah, where she leads research at the intersection of Human-Computer Interaction and Computer Vision. Her work focuses on Computational User Interface Understanding, with specific expertise in applying multimodal generative AI to model human viewing behavior and generate adaptive, accessible UIs. Yue’s research has been published in top venues including CHI, UIST, and CVPR. She initiated and served as the lead organizer of five Computational UI workshops at CHI (2022–2026) and co-organized the Mobile Eye-based Interaction Workshop at ETRA. She was the Accessibility Chair for CHI 2023 and CHI 2024 and will be the Workshop Chair for IUI 2027. Her contributions have been recognized with the Meta Research PhD Fellowship, the Google Europe Scholarship, and the Nokia Scholarship. She was also selected as a Heidelberg Laureate Forum Young Researcher and a UCSD Data Science Rising Star. Previously, Yue has worked or interned at Intel, Adobe Research, Apple, the Max Planck Institute, and Carnegie Mellon University.
Zoya Bylinskii, Adobe Firefly
Perceptual Insights for GenAI Evaluation
In just a few years, visual generative AI models have advanced from struggling to produce the right body morphology to generating portraits indistinguishable from professional headshots. Contributing to this rapid progress are advances in model evaluation—the large-scale measurement of model quality, realism, and harm. Evaluation methods draw from both crowdsourced annotation and perceptual science toolboxes: evaluators may be asked to detect visual artifacts, assess realism, classify content, compare sets, or assign quality ratings. In perceptual science terms, they are effectively performing visual search, categorization, ensemble and memory tasks among others. In this talk, I’ll share insights from developing the evaluation strategy for Adobe Firefly and explore how evaluation methods have been evolving alongside model capabilities. I’ll argue that the deep body of knowledge on human perception and experimental design remains underutilized—and that grounding GenAI evaluation in perceptual science can yield richer, more reliable measures of model quality and usefulness. Zoya Bylinskii leads the Scientific Evaluation Team for Adobe Firefly's suite of generative imaging and video models. Previously, she was a Senior Research Scientist in the Imaging & Languages Lab at Adobe Research. She was an Executive Board Member of The Readability Consortium, an affiliate at MIT and the Harvard Institute of Applied Computational Science. Zoya received her Computer Science Ph.D. from MIT in 2018, and an Hon. B.Sc. in Computer Science and Statistics from the University of Toronto in 2012. Zoya was a 2023 Adobe Founder's Award Recipient, 2018 EECS Rising Star, 2016 Adobe Research Fellow, 2014-2016 NSERC Postgraduate Scholar, 2013 Julie Payette Research Scholar, and 2011 Anita Borg Scholar. Zoya's expertise lies at the interface of human perception & cognition, machine learning, and human-computer interaction - with applications to photography, graphic design, readability, and generative AI.
Perceptual Insights for GenAI Evaluation
In just a few years, visual generative AI models have advanced from struggling to produce the right body morphology to generating portraits indistinguishable from professional headshots. Contributing to this rapid progress are advances in model evaluation—the large-scale measurement of model quality, realism, and harm. Evaluation methods draw from both crowdsourced annotation and perceptual science toolboxes: evaluators may be asked to detect visual artifacts, assess realism, classify content, compare sets, or assign quality ratings. In perceptual science terms, they are effectively performing visual search, categorization, ensemble and memory tasks among others. In this talk, I’ll share insights from developing the evaluation strategy for Adobe Firefly and explore how evaluation methods have been evolving alongside model capabilities. I’ll argue that the deep body of knowledge on human perception and experimental design remains underutilized—and that grounding GenAI evaluation in perceptual science can yield richer, more reliable measures of model quality and usefulness. Zoya Bylinskii leads the Scientific Evaluation Team for Adobe Firefly's suite of generative imaging and video models. Previously, she was a Senior Research Scientist in the Imaging & Languages Lab at Adobe Research. She was an Executive Board Member of The Readability Consortium, an affiliate at MIT and the Harvard Institute of Applied Computational Science. Zoya received her Computer Science Ph.D. from MIT in 2018, and an Hon. B.Sc. in Computer Science and Statistics from the University of Toronto in 2012. Zoya was a 2023 Adobe Founder's Award Recipient, 2018 EECS Rising Star, 2016 Adobe Research Fellow, 2014-2016 NSERC Postgraduate Scholar, 2013 Julie Payette Research Scholar, and 2011 Anita Borg Scholar. Zoya's expertise lies at the interface of human perception & cognition, machine learning, and human-computer interaction - with applications to photography, graphic design, readability, and generative AI.
Vaibhav Unhelkar, Rice University
Training Humans to Work with Robots
We are steadily approaching a future where humans work with robotic assistants, teammates, and even tutors. While significant effort is being dedicated to training robots to work with humans, much less attention has been given to the human side of this partnership. For instance, we see increasingly capable AI models and algorithms that enable robots to infer human intent, plan interactions, and communicate with them. However, human users often lack understanding of how robots will react in novel situations and receive minimal training to work with robots. To ensure the safe and effective use of robots, I argue that training human users is equally critical. This talk will present human-centered computing methods for generating and delivering such training. First, I will discuss Explainable AI methods that can help generate the training content, enabling humans to develop a Theory of Mind regarding the robot. Next, I will discuss mechanisms for delivering this training interactively using physiological sensors (such as eye trackers) and interactive simulation. The talk will conclude with an overview of open problems in making this training more effective through personalization. Vaibhav Unhelkar is an Assistant Professor at Rice University, where he leads the Human-Centered AI & Robotics group. His work spans the development of robotic assistants, intelligent tutors, and decision-support systems aimed at enhancing human performance in domains ranging from healthcare to disaster response. Underpinning these systems are Unhelkar's contributions to imitation learning and explainable AI, designed to train robots, humans, and human-AI teams. He earned his Ph.D. in Autonomous Systems from MIT in 2020 and completed his undergraduate studies at IIT Bombay in 2012. Unhelkar is actively engaged in the AI and robotics research communities, serving on the editorial board of robotics and AI conferences. His research and service have been recognized with awards from the International Foundation for Autonomous Agents and Multi-Agent Systems (AAMAS), among others.
Training Humans to Work with Robots
We are steadily approaching a future where humans work with robotic assistants, teammates, and even tutors. While significant effort is being dedicated to training robots to work with humans, much less attention has been given to the human side of this partnership. For instance, we see increasingly capable AI models and algorithms that enable robots to infer human intent, plan interactions, and communicate with them. However, human users often lack understanding of how robots will react in novel situations and receive minimal training to work with robots. To ensure the safe and effective use of robots, I argue that training human users is equally critical. This talk will present human-centered computing methods for generating and delivering such training. First, I will discuss Explainable AI methods that can help generate the training content, enabling humans to develop a Theory of Mind regarding the robot. Next, I will discuss mechanisms for delivering this training interactively using physiological sensors (such as eye trackers) and interactive simulation. The talk will conclude with an overview of open problems in making this training more effective through personalization. Vaibhav Unhelkar is an Assistant Professor at Rice University, where he leads the Human-Centered AI & Robotics group. His work spans the development of robotic assistants, intelligent tutors, and decision-support systems aimed at enhancing human performance in domains ranging from healthcare to disaster response. Underpinning these systems are Unhelkar's contributions to imitation learning and explainable AI, designed to train robots, humans, and human-AI teams. He earned his Ph.D. in Autonomous Systems from MIT in 2020 and completed his undergraduate studies at IIT Bombay in 2012. Unhelkar is actively engaged in the AI and robotics research communities, serving on the editorial board of robotics and AI conferences. His research and service have been recognized with awards from the International Foundation for Autonomous Agents and Multi-Agent Systems (AAMAS), among others.
Organizers
Prof. Dr. Arantxa Villanueva
Arantzazu Villanueva Larre is currently a full professor in the area of Signal Theory and Communications within the Department of Electrical, Electronic, and Communication Engineering at the Public University of Navarre. She specializes in Signal Theory and Communications and has taught courses on image processing, computer vision, and multimedia signal processing at both undergraduate and master’s levels. Her main research focus is on image-based eye-tracking systems as a tool for human-computer interaction. She also conducts research in the field of medical image analysis. She has served as a co-chair for several conferences, such as the Communication by Gaze Interaction (COGAIN) conference in 2009 and the European Conference on Eye Movements (ECEM) in 2013. Additionally, she has held roles such as Poster Chair at various ETRA conferences and Doctoral Consortium Chair. Her involvement also extends to organizing workshops, such as the "How an Eyetracker Works – Looking Inside the Black Box" workshop, which was held alongside Jeff Pelz and Dixon Cleveland at ECEM 2013. She also contributed to a workshop on IR Eye Safety in Brussels in 2008.
Public University of Navarra, Spain
(avilla@unavarra.es)Arantzazu Villanueva Larre is currently a full professor in the area of Signal Theory and Communications within the Department of Electrical, Electronic, and Communication Engineering at the Public University of Navarre. She specializes in Signal Theory and Communications and has taught courses on image processing, computer vision, and multimedia signal processing at both undergraduate and master’s levels. Her main research focus is on image-based eye-tracking systems as a tool for human-computer interaction. She also conducts research in the field of medical image analysis. She has served as a co-chair for several conferences, such as the Communication by Gaze Interaction (COGAIN) conference in 2009 and the European Conference on Eye Movements (ECEM) in 2013. Additionally, she has held roles such as Poster Chair at various ETRA conferences and Doctoral Consortium Chair. Her involvement also extends to organizing workshops, such as the "How an Eyetracker Works – Looking Inside the Black Box" workshop, which was held alongside Jeff Pelz and Dixon Cleveland at ECEM 2013. She also contributed to a workshop on IR Eye Safety in Brussels in 2008.
Prof. Dr. Enkelejda Kasneci
Enkelejda Kasneci is a Distinguished Professor (“Liesel Beckmann Distinguished Professorship”) for Human-Centered Technologies for Learning at the School of Social Sciences & Technology and in her second affiliation at the School of Computation, Information and Technology. Her research focuses on human-centered technologies, emphasizing the crossroads between multimodal interaction and cutting-edge technological tools like VR, AR, and eye-tracking methodologies. She is member of the ETRA Steering Committee and served as general co-chair of ETRA 2022 and ETRA 2023.
Technical University of Munich, Germany
(enkelejda.kasneci@tum.de)Enkelejda Kasneci is a Distinguished Professor (“Liesel Beckmann Distinguished Professorship”) for Human-Centered Technologies for Learning at the School of Social Sciences & Technology and in her second affiliation at the School of Computation, Information and Technology. Her research focuses on human-centered technologies, emphasizing the crossroads between multimodal interaction and cutting-edge technological tools like VR, AR, and eye-tracking methodologies. She is member of the ETRA Steering Committee and served as general co-chair of ETRA 2022 and ETRA 2023.
Prof. Dr. Gjergji Kasneci
Gjergji Kasneci is a full professor of Responsible Data Science at the Technical University of Munich and a core member of the Munich Data Science Institute. He served as the Chief Technology Officer from 2017 to 2022 at SCHUFA Holding AG, and an Honorary Professor at the University of Tübingen from 2018 until 2023. His research focuses on transparency, robustness, bias, and fairness in machine learning algorithms and involves ethical, legal, and societal considerations with the goal of using artificial intelligence responsibly for the benefit of individuals and society. He is a Fellow of the Konrad Zuse School for Reliable AI and serves on several boards and program committees of renowned conferences, including AAAI, NeurIPS, xAI, DSP, and many more.
Technical University of Munich, Germany
(gjergji.kasneci@tum.de)Gjergji Kasneci is a full professor of Responsible Data Science at the Technical University of Munich and a core member of the Munich Data Science Institute. He served as the Chief Technology Officer from 2017 to 2022 at SCHUFA Holding AG, and an Honorary Professor at the University of Tübingen from 2018 until 2023. His research focuses on transparency, robustness, bias, and fairness in machine learning algorithms and involves ethical, legal, and societal considerations with the goal of using artificial intelligence responsibly for the benefit of individuals and society. He is a Fellow of the Konrad Zuse School for Reliable AI and serves on several boards and program committees of renowned conferences, including AAAI, NeurIPS, xAI, DSP, and many more.
Prof. Dr. Yusuke Sugano
Yusuke Sugano is an associate professor at the Institute of Industrial Science at the University of Tokyo. He was previously an associate professor at the Graduate School of Information Science and Technology, Osaka University, a postdoctoral researcher at the Max Planck Institute for Informatics, and a project research associate at the Institute of Industrial Science, the University of Tokyo. His research interests focus on computer vision and human-computer interaction. He has served as a General Chair for ETRA.
University of Tokyo, Japan
(sugano@iis.u-tokyo.ac.jp)Yusuke Sugano is an associate professor at the Institute of Industrial Science at the University of Tokyo. He was previously an associate professor at the Graduate School of Information Science and Technology, Osaka University, a postdoctoral researcher at the Max Planck Institute for Informatics, and a project research associate at the Institute of Industrial Science, the University of Tokyo. His research interests focus on computer vision and human-computer interaction. He has served as a General Chair for ETRA.
Dr. Yao Rong
Yao Rong is a Junior Fellow at the Rice Academy of Fellows working in the Computer Science Department. She earned her Ph.D. in Computer Science from the Technical University of Munich in 2024. Her research focuses on integrating human factors into AI model design to enhance user experience in human-AI interactions and advancing the trustworthiness of AI systems through the development of human-centered explainable AI techniques. She served as the Chair for Diversity and Inclusion from 2022 to 2024. For instance, during ETRA 2023, she organized a workshop as part of a Diversity Event.
Rice University, United States
(yao.rong@rice.edu)Yao Rong is a Junior Fellow at the Rice Academy of Fellows working in the Computer Science Department. She earned her Ph.D. in Computer Science from the Technical University of Munich in 2024. Her research focuses on integrating human factors into AI model design to enhance user experience in human-AI interactions and advancing the trustworthiness of AI systems through the development of human-centered explainable AI techniques. She served as the Chair for Diversity and Inclusion from 2022 to 2024. For instance, during ETRA 2023, she organized a workshop as part of a Diversity Event.