Raúl Montoliu
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Raúl Montoliu has a total of 111 co-authors. Here are the top co-authors by number of shared publications:
- Joaquín Torres-Sospedra
Universitat de València
12 shared publications - Óscar Belmonte-Fernández
Department of Computer Languages and Systems, Institute of New Imaging Technologies, University Jaume I, Castellón de la Plana, Spain
10 shared publications - Emilio Sansano-Sansano
Universitat Jaume I
9 shared publications - Alex Martínez-Martínez
Universitat Jaume I
6 shared publications Filiberto Pla
Jaume I University
6 shared publications- Inmaculada Remolar
Institute of New Imaging Technologies, Universitat Jaume I, Castellón de la Plana, Spain
6 shared publications
Latest Publications
- Open Access
- 5 November 2026
- Published by Springer Nature in Applied Intelligence
- Vol. 56 (16)
A first step toward structured educational recommendations: knowledge graph embeddings in partially ordered graphs
- 5 November 2026
- Published by Springer Nature in Applied Intelligence
- Vol. 56 (16)
Open AccessA first step toward structured educational recommendations: knowledge graph embeddings in partially ordered graphs
- Open Access
Computing a Characteristic Orientation for Rotation-Independent Image Analysis
Handling geometric transformations, particularly rotations, remains a challenge in deep learning for computer vision. Standard neural networks lack inherent rotation invariance and typically rely on data augmentation or architectural modifications to improve robustness. Although effective, these approaches increase computational demands, require specialised implementations, or alter network structures, limiting their applicability. This paper introduces General Intensity Direction (GID), a preprocessing method that improves rotation robustness without modifying the network architecture. The method estimates a global orientation for each image and aligns it to a canonical reference frame, allowing standard models to process inputs more consistently across different rotations. Unlike moment-based approaches that extract invariant descriptors, this method directly transforms the image while preserving spatial structure, making it compatible with convolutional networks. Experimental evaluation on the rotated MNIST dataset shows that the proposed method achieves higher accuracy than state-of-the-art rotation-invariant architectures. Additional experiments on the CIFAR-10 dataset, confirm that the method remains effective under more complex conditions.
Open AccessComputing a Characteristic Orientation for Rotation-Independent Image Analysis
Handling geometric transformations, particularly rotations, remains a challenge in deep learning for computer vision. Standard neural networks lack inherent rotation invariance and typically rely on data augmentation or architectural modifications to improve robustness. Although effective, these approaches increase computational demands, require specialised implementations, or alter network structures, limiting their applicability. This paper introduces General Intensity Direction (GID), a preprocessing method that improves rotation robustness without modifying the network architecture. The method estimates a global orientation for each image and aligns it to a canonical reference frame, allowing standard models to process inputs more consistently across different rotations. Unlike moment-based approaches that extract invariant descriptors, this method directly transforms the image while preserving spatial structure, making it compatible with convolutional networks. Experimental evaluation on the rotated MNIST dataset shows that the proposed method achieves higher accuracy than state-of-the-art rotation-invariant architectures. Additional experiments on the CIFAR-10 dataset, confirm that the method remains effective under more complex conditions.
- Open Access
- 15 January 2026
- Published by Springer Nature in Scientific Data
- Vol. 13 (1)
RAID-Dataset: human responses to affine image distortions and Gaussian noise
Image quality datasets are used to train and evaluate predictive models of subjective human perception. However, most existing datasets focus on distortions commonly found in digital media and not in natural conditions. Affine transformations are particularly relevant for study, as they are among the most commonly encountered by human observers in everyday life. This Data Descriptor presents a set of human responses to suprathreshold affine image transformations (rotation, translation, scaling) and Gaussian noise as convenient reference to compare with existing image quality datasets. The responses were measured using well-established psychophysics: the Maximum Likelihood Difference Scaling (MLDS). The set contains responses to 864 distorted images. The experiments involved 210 observers and over 40,000 image quadruple comparisons. The dataset is validated by two facts: (a) the responses reproduce classical absolute detection thresholds of the affine and Gaussian distortions, and (b) the responses to Gaussian distortion are correlated to the Mean Opinion Score (MOS) of conventional image quality databases for that distortion. Moreover, the classical Piéron’s law applies to the reaction times of the dataset, and Group-MAD adversarial stimuli reveal that MLDS perceptual scales are more accurate than the conventional MOS.
- 15 January 2026
- Published by Springer Nature in Scientific Data
- Vol. 13 (1)
Open AccessRAID-Dataset: human responses to affine image distortions and Gaussian noise
Image quality datasets are used to train and evaluate predictive models of subjective human perception. However, most existing datasets focus on distortions commonly found in digital media and not in natural conditions. Affine transformations are particularly relevant for study, as they are among the most commonly encountered by human observers in everyday life. This Data Descriptor presents a set of human responses to suprathreshold affine image transformations (rotation, translation, scaling) and Gaussian noise as convenient reference to compare with existing image quality datasets. The responses were measured using well-established psychophysics: the Maximum Likelihood Difference Scaling (MLDS). The set contains responses to 864 distorted images. The experiments involved 210 observers and over 40,000 image quadruple comparisons. The dataset is validated by two facts: (a) the responses reproduce classical absolute detection thresholds of the affine and Gaussian distortions, and (b) the responses to Gaussian distortion are correlated to the Mean Opinion Score (MOS) of conventional image quality databases for that distortion. Moreover, the classical Piéron’s law applies to the reaction times of the dataset, and Group-MAD adversarial stimuli reveal that MLDS perceptual scales are more accurate than the conventional MOS.
- Open Access
- 14 January 2026
- Published by Physical Education Pedagogues Association, Ljubuski in Acta kinesiologica
- p. 34-42
Ecological Validity of Maximal Exercise Tests to Simulate Competitive Demands in Amateur Female Handball
Purpose: Laboratory tests are fundamental for assessing handball performance, but their ecological validity for amateurfemale players is unclear. This exploratory study compared physiological and perceptual responses between commonmaximal exercise tests and official match play to determine if they replicate competitive demands.Methods: Sixteen amateur female handball players completed three laboratory tests (Wingate, cycle ergometer, treadmill)and were monitored during ten official matches. The data collection included both physiological and psychological indicators:heart rate (HR), blood lactate concentration, rating of perceived exertion (RPE), perceived stress, and mood states.Results: Peak HR values recorded in the Wingate and Cycle ergometer were lower than those reached during actualcompetition (d = -.77 to -.75, P< .05). In addition, compared with the Wingate and cycle ergometer tests, competitive matchesrequired players to spend a significantly larger proportion of time with their HR exceeding 90% of its maximum. Thiscontrast was less evident when compared with the treadmill test. A clear mismatch was observed between physiological andperceptual measures. Despite showing stronger physiological strain during matches -such as greater weight loss and higherpre-exercise lactate levels- players reported significantly lower RPE values than in the laboratory conditions (d = .27 to .36,P< .05).Conclusions: None of the laboratory-based tests accurately reproduced the full demands of a handball match. Althoughthe treadmill test elicited similar peak HR values, it failed to reflect the sport’s intermittent nature. The Wingate and cycletests showed even greater discrepancies in cardiovascular intensity. These findings underline the importance of developingsport-specific or hybrid assessment protocols that integrate intermittent workloads and cognitive demands for a more validevaluation of performance.
- 14 January 2026
- Published by Physical Education Pedagogues Association, Ljubuski in Acta kinesiologica
- p. 34-42
Open AccessEcological Validity of Maximal Exercise Tests to Simulate Competitive Demands in Amateur Female Handball
Purpose: Laboratory tests are fundamental for assessing handball performance, but their ecological validity for amateurfemale players is unclear. This exploratory study compared physiological and perceptual responses between commonmaximal exercise tests and official match play to determine if they replicate competitive demands.Methods: Sixteen amateur female handball players completed three laboratory tests (Wingate, cycle ergometer, treadmill)and were monitored during ten official matches. The data collection included both physiological and psychological indicators:heart rate (HR), blood lactate concentration, rating of perceived exertion (RPE), perceived stress, and mood states.Results: Peak HR values recorded in the Wingate and Cycle ergometer were lower than those reached during actualcompetition (d = -.77 to -.75, P< .05). In addition, compared with the Wingate and cycle ergometer tests, competitive matchesrequired players to spend a significantly larger proportion of time with their HR exceeding 90% of its maximum. Thiscontrast was less evident when compared with the treadmill test. A clear mismatch was observed between physiological andperceptual measures. Despite showing stronger physiological strain during matches -such as greater weight loss and higherpre-exercise lactate levels- players reported significantly lower RPE values than in the laboratory conditions (d = .27 to .36,P< .05).Conclusions: None of the laboratory-based tests accurately reproduced the full demands of a handball match. Althoughthe treadmill test elicited similar peak HR values, it failed to reflect the sport’s intermittent nature. The Wingate and cycletests showed even greater discrepancies in cardiovascular intensity. These findings underline the importance of developingsport-specific or hybrid assessment protocols that integrate intermittent workloads and cognitive demands for a more validevaluation of performance.
- Open Access
- 14 October 2025
- Published by MDPI AG in Big Data and Cognitive Computing
- Vol. 9 (10) , 259
Towards the Adoption of Recommender Systems in Online Education: A Framework and Implementation
The rapid expansion of online education has generated large volumes of learner interaction data, highlighting the need for intelligent systems capable of transforming this information into personalized guidance. Educational Recommender Systems (ERS) represent a key application of big data analytics and machine learning, offering adaptive learning pathways that respond to diverse student needs. For widespread adoption, these systems must align with pedagogical principles while ensuring transparency, interpretability, and seamless integration into Learning Management Systems (LMS). This paper introduces a comprehensive framework and implementation of an ERS designed for platforms such as Moodle. The system integrates big data processing pipelines to support scalability, real-time interaction, and multi-layered personalization, including data collection, preprocessing, recommendation generation, and retrieval. A detailed use case demonstrates its deployment in a real educational environment, underlining both technical feasibility and pedagogical value. Finally, the paper discusses challenges such as data sparsity, learner model complexity, and evaluation of effectiveness, offering directions for future research at the intersection of big data technologies and digital education. By bridging theoretical models with operational platforms, this work contributes to sustainable and data-driven personalization in online learning ecosystems.
- 14 October 2025
- Published by MDPI AG in Big Data and Cognitive Computing
- Vol. 9 (10) , 259
Open AccessTowards the Adoption of Recommender Systems in Online Education: A Framework and Implementation
The rapid expansion of online education has generated large volumes of learner interaction data, highlighting the need for intelligent systems capable of transforming this information into personalized guidance. Educational Recommender Systems (ERS) represent a key application of big data analytics and machine learning, offering adaptive learning pathways that respond to diverse student needs. For widespread adoption, these systems must align with pedagogical principles while ensuring transparency, interpretability, and seamless integration into Learning Management Systems (LMS). This paper introduces a comprehensive framework and implementation of an ERS designed for platforms such as Moodle. The system integrates big data processing pipelines to support scalability, real-time interaction, and multi-layered personalization, including data collection, preprocessing, recommendation generation, and retrieval. A detailed use case demonstrates its deployment in a real educational environment, underlining both technical feasibility and pedagogical value. Finally, the paper discusses challenges such as data sparsity, learner model complexity, and evaluation of effectiveness, offering directions for future research at the intersection of big data technologies and digital education. By bridging theoretical models with operational platforms, this work contributes to sustainable and data-driven personalization in online learning ecosystems.
- Open Access
- 16 May 2025
- Published by Elsevier in Internet of Things
- Vol. 32
Time matters: Empirical insights into the limits and challenges of temporal generalization in CSI-based Wi-Fi sensing
- 16 May 2025
- Published by Elsevier in Internet of Things
- Vol. 32
Open AccessTime matters: Empirical insights into the limits and challenges of temporal generalization in CSI-based Wi-Fi sensing
- Open Access
Extracting Daily Routines from Raw RSSI Data
Detecting behavioral routines is an important research area with many implications in various practical applications. One such application involves studying the behavior of older adults residing in care homes. This paper proposes a comprehensive methodology for extracting and analyzing the daily routines of older adults in care homes. The methodology utilizes raw data comprising signal strength measurements obtained from smartwatches worn by six volunteers over five months. To establish the basis for estimating daily activities, fingerprint-based localization techniques are employed to track the minute-by-minute location of each volunteer. Subsequently, the activity performed by each volunteer is estimated for each day. Finally, the study estimates the probability of a user undertaking each one of the studied activities on a given weekday.
Open AccessExtracting Daily Routines from Raw RSSI Data
Detecting behavioral routines is an important research area with many implications in various practical applications. One such application involves studying the behavior of older adults residing in care homes. This paper proposes a comprehensive methodology for extracting and analyzing the daily routines of older adults in care homes. The methodology utilizes raw data comprising signal strength measurements obtained from smartwatches worn by six volunteers over five months. To establish the basis for estimating daily activities, fingerprint-based localization techniques are employed to track the minute-by-minute location of each volunteer. Subsequently, the activity performed by each volunteer is estimated for each day. Finally, the study estimates the probability of a user undertaking each one of the studied activities on a given weekday.
- Open Access
- 6 November 2024
- Published by Elsevier in Data in Brief
- Vol. 57, 111116
A new dataset for spatial, temporal, and tactical analysis of female handball player movements
- 6 November 2024
- Published by Elsevier in Data in Brief
- Vol. 57, 111116
Open AccessA new dataset for spatial, temporal, and tactical analysis of female handball player movements
- Open Access
Which videos are better for the students? Analyzing the student behavior and video metadata
- 1 March 2024
- Published by IATED Academy in INTED Proceedings
- Vol. 1, 4056-4064
ADVANCING VIRTUAL LEARNING: A REVIEW OF MOODLE FOR THE OPTIMIZATION OF ONLINE EDUCATION
- 1 March 2024
- Published by IATED Academy in INTED Proceedings
- Vol. 1, 4056-4064
ADVANCING VIRTUAL LEARNING: A REVIEW OF MOODLE FOR THE OPTIMIZATION OF ONLINE EDUCATION