Ramon Mas-Sansó
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Ramon Mas-Sansó has a total of 18 co-authors. Here are the top co-authors by number of shared publications:
- Maria Francesca Roig-Maimó
University of the Balearic Islands
21 shared publications - Ian Scott MacKenzie
Department of Electrical Engineering and Computer Science, York University, Toronto, Ontario, Canada
9 shared publications - Iosune Salinas-Bueno
University of the Balearic Islands
8 shared publications - Esperança Amengual
Universitat de les Illes Balears
6 shared publications - Miquel Mascaró-Oliver
Universitat de les Illes Balears
6 shared publications - Javier Varona
Department of Mathematics and Computer Science, University of the Balearic Islands, 07122 Palma, Spain
4 shared publications
Latest Publications
- 19 July 2026
- Published by Springer Nature
- p. 165-176
Does Touchscreen Position on a Social Robot Impact Usability?
- 19 July 2026
- Published by Springer Nature
- p. 165-176
Does Touchscreen Position on a Social Robot Impact Usability?
- 7 July 2026
- Published by Springer Nature
- p. 324-340
To Show or Not To Show? Adding a Center Mark for Eye Tracking Target Selection
- 7 July 2026
- Published by Springer Nature
- p. 324-340
To Show or Not To Show? Adding a Center Mark for Eye Tracking Target Selection
- Open Access
- 9 June 2026
- Published by Association for Computing Machinery (ACM)
- p. 1-7
Gazepaths: Using Gaze-based Target-selection Tasks for Tuning Parameters in Scanpath Algorithms
- 9 June 2026
- Published by Association for Computing Machinery (ACM)
- p. 1-7
Open AccessGazepaths: Using Gaze-based Target-selection Tasks for Tuning Parameters in Scanpath Algorithms
- Open Access
- 31 May 2026
- Published by Association for Computing Machinery (ACM)
- p. 1-6
A Standardized Comparison of Blink, Wink, and Dwell as Selection Methods for Gaze-based Interaction
Gaze-based interaction is increasingly used in hands-free and accessible interfaces. Although modern eye-trackers provide sufficient spatial and temporal resolution, accurate and efficient target selection remains a challenge. In this work, we compared three selection methods (wink, blink, dwell) and tested whether displaying or hiding the cursor affects performance in a 2D pointing task. Testing was standardized, as per ISO 9241-411, with performance evaluated using throughput, movement time, and error rate. Cursor visibility produced no significant effects. Concerning throughput, blink achieved the highest value (2.11 bps), followed by dwell-time (1.95 bps; ≈ 7.5% lower), whereas wink performed worst (1.56 bps). Although throughput was higher, only 16.7% of participants preferred blink; most favored dwell-time. Error rates further disadvantaged wink (20.4%), while dwell-time showed the lowest errors (1.82%). Overall, considering throughput, error rate, and user preference, wink is not recommended. Dwell-time selection is the best overall choice, offering a better balance for gaze-controlled interaction.
- 31 May 2026
- Published by Association for Computing Machinery (ACM)
- p. 1-6
Open AccessA Standardized Comparison of Blink, Wink, and Dwell as Selection Methods for Gaze-based Interaction
Gaze-based interaction is increasingly used in hands-free and accessible interfaces. Although modern eye-trackers provide sufficient spatial and temporal resolution, accurate and efficient target selection remains a challenge. In this work, we compared three selection methods (wink, blink, dwell) and tested whether displaying or hiding the cursor affects performance in a 2D pointing task. Testing was standardized, as per ISO 9241-411, with performance evaluated using throughput, movement time, and error rate. Cursor visibility produced no significant effects. Concerning throughput, blink achieved the highest value (2.11 bps), followed by dwell-time (1.95 bps; ≈ 7.5% lower), whereas wink performed worst (1.56 bps). Although throughput was higher, only 16.7% of participants preferred blink; most favored dwell-time. Error rates further disadvantaged wink (20.4%), while dwell-time showed the lowest errors (1.82%). Overall, considering throughput, error rate, and user preference, wink is not recommended. Dwell-time selection is the best overall choice, offering a better balance for gaze-controlled interaction.
- Open Access
- 23 December 2025
- Published by Asociación Interacción Persona-Ordenador in Revista de la Asociación Interacción Persona Ordenador (AIPO)
- Vol. 6 (2) , 58-69
UIBAIFED: Un dataset de expresiones faciales generado por IA para visibilizar la diversidad
Este artículo presenta UIBAIFED, un nuevo conjunto de datos de expresiones faciales compuesto por imágenes realistas y de alta calidad, etiquetadas con grupo de edad, género, etnia y 22 microexpresiones basadas en las expresiones universales de Ekman y la taxonomía de Gary Faigin. El conjunto de datos fue generado mediante modelos de difusión y está diseñado para mejorar la investigación en reconocimiento de expresiones faciales, aumentado la representación étnica, generacional y de complexión física, visibilizando la diversidad existente en el mundo real. La validación inicial mediante una red neuronal convolucional (CNN) alcanzó una precisión del 80% en la clasificación de microexpresiones. Además, se realizó un experimento con participantes humanos con un subconjunto de datos formado por imágenes del grupo de edad de personas mayores (85 años), colectivo tradicionalmente marginado en los conjuntos de datos existentes. Los mejores resultados de clasificación fueron para las emociones de tristeza (94.6%), neutral (92.8%) y alegría (83.8%), mientras que las peores emociones identificadas fueron las emociones de miedo (33.9%) y asco (34.7%), con una precisión global del 74%. UIBAIFED ofrece un nivel de etiquetado más detallado que los conjuntos de datos existentes, lo que facilita el análisis del rendimiento en sistemas de reconocimiento de expresiones faciales en distintos grupos demográficos y contribuye al desarrollo de modelos más robustos y generalizables.
- 23 December 2025
- Published by Asociación Interacción Persona-Ordenador in Revista de la Asociación Interacción Persona Ordenador (AIPO)
- Vol. 6 (2) , 58-69
Open AccessUIBAIFED: Un dataset de expresiones faciales generado por IA para visibilizar la diversidad
Este artículo presenta UIBAIFED, un nuevo conjunto de datos de expresiones faciales compuesto por imágenes realistas y de alta calidad, etiquetadas con grupo de edad, género, etnia y 22 microexpresiones basadas en las expresiones universales de Ekman y la taxonomía de Gary Faigin. El conjunto de datos fue generado mediante modelos de difusión y está diseñado para mejorar la investigación en reconocimiento de expresiones faciales, aumentado la representación étnica, generacional y de complexión física, visibilizando la diversidad existente en el mundo real. La validación inicial mediante una red neuronal convolucional (CNN) alcanzó una precisión del 80% en la clasificación de microexpresiones. Además, se realizó un experimento con participantes humanos con un subconjunto de datos formado por imágenes del grupo de edad de personas mayores (85 años), colectivo tradicionalmente marginado en los conjuntos de datos existentes. Los mejores resultados de clasificación fueron para las emociones de tristeza (94.6%), neutral (92.8%) y alegría (83.8%), mientras que las peores emociones identificadas fueron las emociones de miedo (33.9%) y asco (34.7%), con una precisión global del 74%. UIBAIFED ofrece un nivel de etiquetado más detallado que los conjuntos de datos existentes, lo que facilita el análisis del rendimiento en sistemas de reconocimiento de expresiones faciales en distintos grupos demográficos y contribuye al desarrollo de modelos más robustos y generalizables.
- Open Access
- 3 November 2025
- Published by Elsevier in Applied Ergonomics
- Vol. 131, 104680
Design of 2D Fitts’ law experiments: An odd thing about targets
- 3 November 2025
- Published by Elsevier in Applied Ergonomics
- Vol. 131, 104680
Open AccessDesign of 2D Fitts’ law experiments: An odd thing about targets
- Open Access
- 21 September 2025
- Published by Association for Computing Machinery (ACM)
- p. 1-7
Tracing: The Forgotten Task of ISO 9241
- 21 September 2025
- Published by Association for Computing Machinery (ACM)
- p. 1-7
Open AccessTracing: The Forgotten Task of ISO 9241
- Open Access
- 21 September 2025
- Published by Association for Computing Machinery (ACM)
- p. 1-5
RehbeCa in Action! A Demo of Innovative Mobile Cervical Rehabilitation
- 21 September 2025
- Published by Association for Computing Machinery (ACM)
- p. 1-5
Open AccessRehbeCa in Action! A Demo of Innovative Mobile Cervical Rehabilitation
- Open Access
- 28 August 2025
- Published by Springer Nature in Universal Access in the Information Society
- Vol. 24 (4) , 3185-3207
The puzzle of RehbeCa: an exergame for assessing cervicalgia
Neck range of motion (ROM) and pain assessment are crucial aspects of cervical examination. The goal is to match a patient’s clinical presentation with an appropriate treatment. We developed and evaluated a mobile puzzle-based exergame for assessing cervicalgia. An exploratory experiment with healthy participants and simulated neck mobility restrictions investigated the relationship between task performance and neck ROM. Insights from this phase led to refinement of the experimental design and apparatus. In a subsequent experiment, the exergame was tested on participants with neck pain to analyze how their neck ROM and pain condition affect task performance. Results indicate that participants suffering from neck pain maintained similar task performance metrics, such as selection rate, but exhibited greater variability in ROM, reflecting their adaptation to pain and mobility limitations. Results also indicate that pain condition significantly influences completion time: Participants with neck pain took longer to complete puzzles than participants with no pain. In the last stage, the data from both experiments were used to feed a machine learning model to test the automatic prediction of cervicalgia assessment parameters. This work advances rehabilitation technologies by integrating tools to automatically adapt and combine exercise and parameters of cervical rehabilitation enabling functional assessment without needing external sensors, distinguishing it from many existing approaches. The system is not only a rehabilitation tool, but also a method for evaluating cervical pain and mobility based on user interaction data.
- 28 August 2025
- Published by Springer Nature in Universal Access in the Information Society
- Vol. 24 (4) , 3185-3207
Open AccessThe puzzle of RehbeCa: an exergame for assessing cervicalgia
Neck range of motion (ROM) and pain assessment are crucial aspects of cervical examination. The goal is to match a patient’s clinical presentation with an appropriate treatment. We developed and evaluated a mobile puzzle-based exergame for assessing cervicalgia. An exploratory experiment with healthy participants and simulated neck mobility restrictions investigated the relationship between task performance and neck ROM. Insights from this phase led to refinement of the experimental design and apparatus. In a subsequent experiment, the exergame was tested on participants with neck pain to analyze how their neck ROM and pain condition affect task performance. Results indicate that participants suffering from neck pain maintained similar task performance metrics, such as selection rate, but exhibited greater variability in ROM, reflecting their adaptation to pain and mobility limitations. Results also indicate that pain condition significantly influences completion time: Participants with neck pain took longer to complete puzzles than participants with no pain. In the last stage, the data from both experiments were used to feed a machine learning model to test the automatic prediction of cervicalgia assessment parameters. This work advances rehabilitation technologies by integrating tools to automatically adapt and combine exercise and parameters of cervical rehabilitation enabling functional assessment without needing external sensors, distinguishing it from many existing approaches. The system is not only a rehabilitation tool, but also a method for evaluating cervical pain and mobility based on user interaction data.
- Open Access
- 25 April 2025
- Published by Association for Computing Machinery (ACM)
- p. 1-3
Empirical Research Methods for Human-Computer Interaction
- 25 April 2025
- Published by Association for Computing Machinery (ACM)
- p. 1-3
Open AccessEmpirical Research Methods for Human-Computer Interaction