Jorge Azorin-Lopez
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Previous Organizations / Institutes
The organizations / institutes from the scholar's other articles in the past ten years
Research Subjects
Jorge Azorin-Lopez has a total of 172 co-authors. Here are the top co-authors by number of shared publications:
- Andres Fuster-Guillo
University of Alicante
59 shared publications - Jose Garcia-Rodriguez
Department of Computer Technology, University of Alicante, 03690 Alicante, Spain
44 shared publications - Marcelo Saval-Calvo
University Institute for Computing Research, University of Alicante, Alicante, Spain
36 shared publications - Nahuel Garcia-D’uRso
University of Alicante
15 shared publications - Mauricio-Andres Zamora-Hernandez
Department of Industrial Engineering, University of Costa Rica, San Pedro de Montes de Oca, San José 11501-2060, Costa Rica
12 shared publications - David Alfaro-Viquez
Department of Industrial Engineering, University of Costa Rica, San Pedro de Montes de Oca, San José 11501-2060, Costa Rica
10 shared publications
Latest Publications
- Open Access
- 26 September 2026
- Published by MDPI AG in Applied Sciences
- Vol. 16 (19) , 9584
Predicting Avian Observation Patterns in Mediterranean Wetlands: Spatiotemporal Deep Learning Fusion of Multi-Source Surveys, Citizen Science, and Autonomous Acoustic Sensing
Protected wetlands in the Mediterranean are vital biodiversity hotspots that face growing pressure from climate change and human activity. Traditional bird monitoring relies on professional field surveys, which, although highly standardized, are resource-constrained and limited in temporal frequency. Here, we present ValWet-Birds, a multi-source spatiotemporal dataset and fusion framework that harmonises and integrates professional counts, eBird citizen science registries, and passive acoustic monitoring (via a BirdNET classifier deployed on a Raspberry Pi 5 node) for three protected wetlands in Alicante, Spain (El Hondo, Santa Pola, and La Mata–Torrevieja), from 2010 to 2025. We normalize incompatible observation protocols into a unified monthly relative observation proportion target across 12 sub-regions and 394 taxa (349 resolved to species level after a taxonomic audit). Spatiotemporal prediction models are implemented using Multi-Layer Perceptron (MLP) and Long Short-Term Memory (LSTM) neural networks. A preliminary evaluation using non-matched test sets suggested a 42.7% reduction in Test Mean Squared Error for the LSTM under multi-source augmentation. Re-evaluating every source condition on an identical held-out set of professional-census observations shows that this benefit does not hold in aggregate: both baselines outperform naive temporal reference predictors by a wide margin, but adding eBird and BirdNET data does not reduce error relative to census-only training overall. The exception is conservation-relevant taxa (a 50-species subset cross-referenced against Annex I of the EU Birds Directive), for which multi-source augmentation reduces LSTM error by 29% and MLP error by 9%, plausibly because these less-common species have sparser census history to draw on. We report this reversal explicitly as a methodological finding in its own right. Finally, we describe the design and interactive user flows of the deployed web visualization platform Avistory, which presents observation summaries and model outputs; it should not be interpreted as a validated population-monitoring or conservation-decision system.
- 26 September 2026
- Published by MDPI AG in Applied Sciences
- Vol. 16 (19) , 9584
Open AccessPredicting Avian Observation Patterns in Mediterranean Wetlands: Spatiotemporal Deep Learning Fusion of Multi-Source Surveys, Citizen Science, and Autonomous Acoustic Sensing
Protected wetlands in the Mediterranean are vital biodiversity hotspots that face growing pressure from climate change and human activity. Traditional bird monitoring relies on professional field surveys, which, although highly standardized, are resource-constrained and limited in temporal frequency. Here, we present ValWet-Birds, a multi-source spatiotemporal dataset and fusion framework that harmonises and integrates professional counts, eBird citizen science registries, and passive acoustic monitoring (via a BirdNET classifier deployed on a Raspberry Pi 5 node) for three protected wetlands in Alicante, Spain (El Hondo, Santa Pola, and La Mata–Torrevieja), from 2010 to 2025. We normalize incompatible observation protocols into a unified monthly relative observation proportion target across 12 sub-regions and 394 taxa (349 resolved to species level after a taxonomic audit). Spatiotemporal prediction models are implemented using Multi-Layer Perceptron (MLP) and Long Short-Term Memory (LSTM) neural networks. A preliminary evaluation using non-matched test sets suggested a 42.7% reduction in Test Mean Squared Error for the LSTM under multi-source augmentation. Re-evaluating every source condition on an identical held-out set of professional-census observations shows that this benefit does not hold in aggregate: both baselines outperform naive temporal reference predictors by a wide margin, but adding eBird and BirdNET data does not reduce error relative to census-only training overall. The exception is conservation-relevant taxa (a 50-species subset cross-referenced against Annex I of the EU Birds Directive), for which multi-source augmentation reduces LSTM error by 29% and MLP error by 9%, plausibly because these less-common species have sparser census history to draw on. We report this reversal explicitly as a methodological finding in its own right. Finally, we describe the design and interactive user flows of the deployed web visualization platform Avistory, which presents observation summaries and model outputs; it should not be interpreted as a validated population-monitoring or conservation-decision system.
- Open Access
Global Dataset of Solar Power Plants: Multidimensional Integration and Analysis
The global expansion of renewable energy has increased the strategic importance of photovoltaic (PV) power plants and the demand for comprehensive, high-quality solar datasets to support energy planning, optimization, and data-driven applications. However, existing datasets are often constrained by limited attribute integration, incomplete information, and insufficient global coverage. This study presents a standardized, reliable, and multidimensional global dataset for photovoltaic power plant analysis, together with a largely reproducible and documented methodology that automates the collection, generation, and integration of heterogeneous solar-related data from multiple sources. Using this methodology, 27 geographic, topographic, logistical, climatic, and power-related attributes were integrated into a unified dataset comprising 58,978 photovoltaic plant records worldwide. Descriptive statistical analyses were performed to characterize the dataset and assess its informational richness and consistency. The results demonstrate that both the proposed methodology and the resulting dataset provide a robust foundation for photovoltaic energy research and decision-making across diverse application domains. By making this resource publicly available, this work facilitates reproducible research and supports the development of advanced analytical, predictive, and optimization models for academic, industrial, and policy-oriented applications.
Open AccessGlobal Dataset of Solar Power Plants: Multidimensional Integration and Analysis
The global expansion of renewable energy has increased the strategic importance of photovoltaic (PV) power plants and the demand for comprehensive, high-quality solar datasets to support energy planning, optimization, and data-driven applications. However, existing datasets are often constrained by limited attribute integration, incomplete information, and insufficient global coverage. This study presents a standardized, reliable, and multidimensional global dataset for photovoltaic power plant analysis, together with a largely reproducible and documented methodology that automates the collection, generation, and integration of heterogeneous solar-related data from multiple sources. Using this methodology, 27 geographic, topographic, logistical, climatic, and power-related attributes were integrated into a unified dataset comprising 58,978 photovoltaic plant records worldwide. Descriptive statistical analyses were performed to characterize the dataset and assess its informational richness and consistency. The results demonstrate that both the proposed methodology and the resulting dataset provide a robust foundation for photovoltaic energy research and decision-making across diverse application domains. By making this resource publicly available, this work facilitates reproducible research and supports the development of advanced analytical, predictive, and optimization models for academic, industrial, and policy-oriented applications.
- Open Access
A Digital Twin Framework for Multimodal Operator-Centered Human–Cobot Collaboration in Assembly Tasks
Current digital twin frameworks focused on human–robot collaboration rarely take into account the sensory degradation of real industrial environments, nor do they integrate the operator as an active agent within the system. This research presents a multimodal digital twin framework for a dual-arm collaborative robot at an assembly station; the system was developed using ROS 2 Jazzy and CoppeliaSim as the simulator. The architecture integrates three main components: the first is a perception layer that captures voice commands using Whisper ASR and the state of the workspace using a hybrid YOLO + ViT visual pipeline, both with per-channel metadata; the second consists of a Confidence-Weighted Late Fusion engine that dynamically adjusts the weight of each modality based on real-time signal quality, so that each fusion decision can be reconstructed from the signals that generated it; and the third component is a Reference Resolver that grounds linguistic intent within the visual context of the scene and in the fusion weights, using a local instance of Llama 3.1 8B that does not transmit audio, transcripts, or images outside the system. The framework was evaluated using 210 iterations distributed across seven degradation conditions of increasing severity, comparing adaptive fusion against a baseline of fixed weights (0.5/0.5). Under clean conditions and under visual degradation of any severity, both configurations achieved 100% accuracy. Under severe auditory degradation (SNR 0 dB), adaptive fusion activated the safety gate and refrained from executing most commands (13.3% accuracy), while the fixed-weight baseline executed more commands (60% accuracy) but made three incorrect object selections; under severe dual degradation, the pattern repeated (13.3% vs. 40%, with five incorrect selections in the baseline). The adaptive system made no grounding errors in the 210 executions, compared to eight in the baseline, substituting incorrect execution with conservative abstention when no modality provided a reliable signal. The implementation, featuring a versioned degradation protocol and a fixed seed, provides a reproducible benchmark for evaluating multimodal fusion strategies in human–cobot interaction.
Open AccessA Digital Twin Framework for Multimodal Operator-Centered Human–Cobot Collaboration in Assembly Tasks
Current digital twin frameworks focused on human–robot collaboration rarely take into account the sensory degradation of real industrial environments, nor do they integrate the operator as an active agent within the system. This research presents a multimodal digital twin framework for a dual-arm collaborative robot at an assembly station; the system was developed using ROS 2 Jazzy and CoppeliaSim as the simulator. The architecture integrates three main components: the first is a perception layer that captures voice commands using Whisper ASR and the state of the workspace using a hybrid YOLO + ViT visual pipeline, both with per-channel metadata; the second consists of a Confidence-Weighted Late Fusion engine that dynamically adjusts the weight of each modality based on real-time signal quality, so that each fusion decision can be reconstructed from the signals that generated it; and the third component is a Reference Resolver that grounds linguistic intent within the visual context of the scene and in the fusion weights, using a local instance of Llama 3.1 8B that does not transmit audio, transcripts, or images outside the system. The framework was evaluated using 210 iterations distributed across seven degradation conditions of increasing severity, comparing adaptive fusion against a baseline of fixed weights (0.5/0.5). Under clean conditions and under visual degradation of any severity, both configurations achieved 100% accuracy. Under severe auditory degradation (SNR 0 dB), adaptive fusion activated the safety gate and refrained from executing most commands (13.3% accuracy), while the fixed-weight baseline executed more commands (60% accuracy) but made three incorrect object selections; under severe dual degradation, the pattern repeated (13.3% vs. 40%, with five incorrect selections in the baseline). The adaptive system made no grounding errors in the 210 executions, compared to eight in the baseline, substituting incorrect execution with conservative abstention when no modality provided a reliable signal. The implementation, featuring a versioned degradation protocol and a fixed seed, provides a reproducible benchmark for evaluating multimodal fusion strategies in human–cobot interaction.
- Open Access
- 27 June 2026
- Published by Elsevier in SSRN Electronic Journal
Informed Machine Learning in Healthcare: A Systematic Review of Knowledge-Driven Deep Learning for Medical Applications
Deep learning (DL) has driven significant advances in healthcare; however, the field still faces important challenges, including the scarcity of high-quality data, the need for reliable and reproducible models, and the requirement for clinically consistent and interpretable predictions. In this context, the integration of prior knowledge into the learning process has emerged as a promising strategy, giving rise to the paradigm of informed machine learning (IML), which combines learning from training data with external domain-specific knowledge. In this paper, we present a systematic review of IML in healthcare, with emphasis on DL methods applied to medical imaging and clinical text analysis. To this end, we extend the IML taxonomy to categorize existing methods according to the type of knowledge representation, such as mathematical models and knowledge graphs, the point of integration within the learning pipeline, and the source of domain knowledge. Overall, this work provides an analysis of the literature that identifies the main forms of integration and emerging trends in IML, demonstrating how prior knowledge can guide learning and improve performance in complex clinical environments.This review serves as a reference for the development of IML methods and lays the foundation for future advances in knowledge-driven healthcare artificial intelligence.
- 27 June 2026
- Published by Elsevier in SSRN Electronic Journal
Open AccessInformed Machine Learning in Healthcare: A Systematic Review of Knowledge-Driven Deep Learning for Medical Applications
Deep learning (DL) has driven significant advances in healthcare; however, the field still faces important challenges, including the scarcity of high-quality data, the need for reliable and reproducible models, and the requirement for clinically consistent and interpretable predictions. In this context, the integration of prior knowledge into the learning process has emerged as a promising strategy, giving rise to the paradigm of informed machine learning (IML), which combines learning from training data with external domain-specific knowledge. In this paper, we present a systematic review of IML in healthcare, with emphasis on DL methods applied to medical imaging and clinical text analysis. To this end, we extend the IML taxonomy to categorize existing methods according to the type of knowledge representation, such as mathematical models and knowledge graphs, the point of integration within the learning pipeline, and the source of domain knowledge. Overall, this work provides an analysis of the literature that identifies the main forms of integration and emerging trends in IML, demonstrating how prior knowledge can guide learning and improve performance in complex clinical environments.This review serves as a reference for the development of IML methods and lays the foundation for future advances in knowledge-driven healthcare artificial intelligence.
- 16 June 2026
- Published by Springer Nature
- p. 596-607
A Multimodal Signal Fusion Framework for Continuous Fatigue Estimation in Human-Centric Environments
- 16 June 2026
- Published by Springer Nature
- p. 596-607
A Multimodal Signal Fusion Framework for Continuous Fatigue Estimation in Human-Centric Environments
- 2 June 2026
- Published by Springer Nature
- p. 572-581
Photovoltaic Power Estimation Based on Meteorological Data Using LSTM and Attention
- 2 June 2026
- Published by Springer Nature
- p. 572-581
Photovoltaic Power Estimation Based on Meteorological Data Using LSTM and Attention
- 2 June 2026
- Published by Springer Nature
- p. 498-508
Characterization of Photovoltaic Power with SOM and Clustering: An Approach Based on Meteorological Data
- 2 June 2026
- Published by Springer Nature
- p. 498-508
Characterization of Photovoltaic Power with SOM and Clustering: An Approach Based on Meteorological Data
- 1 April 2026
- Published by Springer Nature
- p. 45-50
Multimodal Contrastive Learning for Clinical Data Alignment via Fat Composition Representations
- 1 April 2026
- Published by Springer Nature
- p. 45-50
Multimodal Contrastive Learning for Clinical Data Alignment via Fat Composition Representations
- 1 April 2026
- Published by Springer Nature
- p. 98-104
Informed Zero-Shot Elasmobranch Classification Using Expert Knowledge and Illustrations
- 1 April 2026
- Published by Springer Nature
- p. 98-104
Informed Zero-Shot Elasmobranch Classification Using Expert Knowledge and Illustrations