2025-09-25: ACM Symposium on Eye Tracking Research and Applications (ETRA) 2025 Trip Report

By -

 


ACM Symposium on Eye Tracking Research & Applications (ETRA) 2025 was an in-person conference held at Miraikan, Tokyo, Japan. The conference took place from May 26 to May 29, 2025. 

ETRA 2025 focuses on all aspects of eye movement research across a wide range of disciplines and brings together computer scientists, engineers, and behavioral scientists to advance eye-tracking research and applications. 

Keynote 1

Dr. Yukie Nagai, Project Professor at The University of Tokyo, delivered the first keynote of ETRA 2025. Her talk “How People See the World: An Embodied Predictive Processing Theory”. Dr. Nagai introduced a neuro-inspired theory of human visual perception based on embodied predictive processing. She explained how sensorimotor learning drives the development of visual perception and attention, providing new insights into the mechanisms underlying how people see and interpret the world.

 

Tutorial Highlight

Dr. Andrew T. Duchowski delivered a tutorial “Gaze Analytics: A Data Science Perspective.” The session covered a wide range of topics, from PsychoPy setup to advanced transition entropy analysis. Participants gained hands-on experience with the complete pipeline, engaging directly with both methodology and stimulus generation. This tutorial gave attendees practical exposure to data science approaches in gaze analytics, reinforcing its value for eye-tracking research.

Papers Session 1

The first paper session featured groundbreaking research covering diverse applications of eye tracking. Presentations included work on Large language Model (LLMs) alignment analysis, laser-based eye tracking for smart glasses, and methods for detecting expertise through gaze patterns. The session demonstrated how eye-tracking research continues to expand its scope, integrating with fields such as artificial intelligence, wearable technology, and cognitive modeling. The variety of topics sparked engaging discussions, highlighting the breadth and depth of current innovations in the community.


Dr. Sampath Jayarathna of NirdsLab from Old Dominion University, Dr. Yasith Jayawardana from Georgia Institute of Technology, and Dr. Gavindya Jayawardana from University of Texas at Austin attended the paper sessions, poster sessions, and keynotes of the conference in person.

Dr. Gavindya, a postdoctoral fellow at The University of Texas at Austin working with Dr. Jacek Gwizdka, presented her research on advancing real-time measures of visual attention through the ambient/focal coefficient K "A Real-Time Approach to Capture Ambient and Focal Attention in Visual Search". Their work introduced a robust parametrization and alternative estimation method, along with two new real-time measures analogous to K.

Building on her broader research interests in Eye-Tracking, Human-Computer Interaction, Human-Information Interaction, Data Science, and Machine Learning, Gavindya’s presentation demonstrated how neuro-physiological evidence and cognitive load detection approaches can be integrated into applied systems. 


Johannes Meyer presented “Ambient Light Robust Eye-Tracking for Smart Glasses Using Laser Feedback Interferometry Sensors with Elongated Laser Beams.” This innovative approach focuses on developing eye-tracking methods that remain robust under varying light conditions, a crucial step toward making smart glasses more reliable in real-world environments. 


Mengdi Wang presented “Iris Style Transfer: Enhancing Iris Recognition with Style Features and Privacy Preservation through Neural Style Transfer.” The work, carried out in collaboration with Dr. Efe Bozkir and Dr. Enkelejda Kasneci, explored how neural style transfer techniques can improve iris recognition systems while simultaneously addressing privacy concerns. This presentation highlighted the potential of combining computer vision methods with privacy-preserving mechanisms, making it a noteworthy contribution to the conference program.


Virmarie Maquiling presented “Zero-Shot Pupil Segmentation with SAM 2: A Case Study of Over 14 Million Images”. This large-scale study, conducted in collaboration with Dr. Sean Anthony Byrne, Dr. Diederick C. Niehorster, Dr. Marco Carminati, and Dr. Enkelejda Kasneci, demonstrated the use of the Segment Anything Model (SAM 2) for robust pupil segmentation. By applying zero-shot learning methods to an extensive dataset of over 14 million images, the work highlighted both the scalability and efficiency of modern machine learning in eye-tracking applications.


Sponsor Exhibitions

The sponsors showcase highlighted cutting-edge developments in eye-tracking technology. Exhibitors demonstrated a range of innovative solutions, including immersive VR setups, wearable eye-tracking devices, and advanced software platforms. Attendees had the chance to interact directly with the technology, exploring applications that bring the future of eye tracking to life. These exhibitions not only showcased sponsor contributions but also emphasized the vital role industry partnerships play in driving forward eye-tracking research and applications.


Keynote 2

Dr. Jean-Marc Odobez, a senior researcher at IDIAP and EPFL, and Head of the Perception and Activity Understanding Group, delivered the second keynote, “Looking Through Their Eyes: Decoding Gaze and Attention in Everyday Life.” Dr. Odobez, a leading expert in multimodal perception systems and co-founder of Eyeware SA, presented his team’s work in the areas of gaze analysis and visual focus of attention. His keynote showcased how gaze data can be decoded to better understand naturalistic interactions, emphasizing both the technical challenges and practical applications of attention modeling in real-world settings.


Poster Session

Poster Session showcased 36 innovative research studies highlighting the role of Data, Machine Learning, and AI in eye-tracking applications. The session featured diverse approaches, ranging from neural networks for gaze prediction to advanced analytics methods for interpreting eye-movement data. The room was filled with lively discussions as researchers exchanged ideas, explained methodologies, and received feedback from peers and experts. 



Pahan Jayarathna presented his first publication at the premier ACM eye-tracking conference. His work, titled “Nocturnal Diabetic Hypoglycemia Detection Using Eye Tracking,” introduced a novel approach to monitoring diabetic hypoglycemia through eye-tracking techniques.

The study opened up a promising line of inquiry into how eye-movement data can serve as non-invasive indicators for health conditions, particularly in detecting hypoglycemia during nighttime. Pahan’s poster presentation drew attention from conference attendees, sparking discussions on the medical applications of eye tracking beyond traditional HCI contexts. His contribution marked an important step in extending the reach of eye-tracking research into healthcare and biomedical domains.

 

Paper Session 2

The second paper session emphasized the importance of methodological rigor in eye-tracking research. Presentations ranged from evaluating detection algorithms to applying LLMs for cognitive processing tasks. These studies highlighted how careful methodological design provides the analytical foundations for deriving meaningful insights from eye movement data. The session demonstrated both technical depth and practical relevance, reinforcing the role of rigorous analysis in advancing the reliability and impact of eye-tracking research.

 

Panel Discussion: Eye Tracking for Accessibility

The final day began with an engaging panel discussion on Eye Tracking for Accessibility. The session brought together experts including Dr. Pawel Kasprowski (Silesian University of Technology), Dr. Diako Mardanbegi (American University of Beirut), Dr. Krzysztof Krejtz (SWPS University of Social Sciences and Humanities), and Dr. Hironobu Takagi (IBM Research, Miraikan). The discussion was led by Dr. Mohamed Khamis (University of Glasgow). Panelists explored how eye-tracking technologies can be leveraged to improve accessibility in everyday contexts, addressing challenges and opportunities for making digital and physical environments more inclusive. Their insights underscored the transformative potential of eye tracking to enhance usability and accessibility for diverse populations.


Workshop Session: GenAI Meets Eye Tracking

The GenAI Workshop focused on the intersection of generative AI and eye-tracking research. The session began with a keynote by Xi Wang on “Decoding Human Behavior Through Gaze Patterns”, which explored how gaze data can be harnessed to understand complex aspects of human behavior.

Following the keynote, the workshop featured contributions from Gjergji Kasneci, Enkelejda Kasneci, Aranxa Villanueva, and Yusuke Sugano. Together, the speakers emphasized how generative AI techniques can advance eye-tracking applications, including cognitive modeling, behavioral prediction, and new opportunities for human-computer interaction.

The workshop was well-attended and highly interactive, bringing together perspectives from academia and industry. It highlighted how AI-driven methods are shaping the future of gaze research and opened discussions about challenges in integrating these tools into practical applications.

 

Papers Session 3

The third paper session highlighted innovative Computer Vision approaches applied to eye-tracking research. Presentations explored a range of groundbreaking topics, including iris style transfer for privacy preservation and zero-shot pupil segmentation with SAM 2 applied to over 14 million images. These studies showcased how modern vision-based techniques can push the boundaries of both data-driven analysis and practical applications in eye tracking.

By combining advanced AI methods with large-scale datasets, the session emphasized the critical role of computer vision in addressing challenges of scalability, accuracy, and privacy within the field. Researchers also demonstrated how these developments could translate into real-world solutions, reinforcing the strong connection between technical innovation and human-centered applications in eye-tracking research.


Social Events

ETRA 2025 wasn’t just about research and presentations—it also provided memorable opportunities for networking and community building. One of the highlights was the conference banquet, held aboard a dinner cruise in Tokyo Bay. Attendees enjoyed an evening of fine Japanese cuisine, sake, and vibrant discussions while taking in breathtaking views of the city skyline.

The social events gave researchers, practitioners, and students a chance to relax, connect, and exchange ideas in a more informal setting. These moments of shared meals and conversations strengthened the sense of community within the eye-tracking research field, ensuring that collaborations extend beyond the technical sessions into long-lasting professional relationships.

 

ETRA 2025 emphasized inclusivity and sustainability even during mealtimes. Attendees were served Halal-friendly, vegan, and regular bento options, ensuring that everyone had accessible and culturally sensitive food choices. The beautifully prepared bento boxes showcased a variety of Japanese flavors while catering to diverse dietary needs.

Networking & coffee breaks, beyond the formal sessions, also fostered an environment for informal yet impactful exchanges. it provided a space where researchers, students, and industry professionals gathered to share ideas, sketch concepts on whiteboards, and even run quick demos on their phones. These animated discussions often sparked new collaborations and innovative research directions.

In conclusion, ETRA 2025 in Tokyo was an inspiring event that showcased the latest advancements in eye-tracking research. The conference fostered meaningful discussions, collaborations, and the exchange of ideas, setting the stage for future developments. As we look ahead, the excitement for ETRA 2026 in Marrakesh promises even more groundbreaking research and opportunities for growth. The experiences gained from this event will undoubtedly shape the future of eye-tracking technology and research.


About the Author:
Lawrence Obiuwevwi is a Ph.D. student in the Department of Computer Science, a graduate research assistant with The Center for Secure and Intelligent Critical Systems (CSICS), and a proud student member The Web Science and Digital Libraries (WS-DL) Research Group, and NirdsLab at Old Dominion University.


Lawrence Obiuwevwi
Graduate Research Assistant
Virginia Modeling, Analysis, & Simulation Center
Department of Computer Science
Old Dominion University, Norfolk, VA 23529
Email: lobiu001@odu.edu
Web : lawobiu.com

Comments

Post a Comment