Persona:
Herrera Caro, Pedro Javier

Cargando...
Foto de perfil
Dirección de correo electrónico
ORCID
0000-0001-8679-6617
Fecha de nacimiento
Proyectos de investigación
Unidades organizativas
Puesto de trabajo
Apellidos
Herrera Caro
Nombre de pila
Pedro Javier
Nombre

Resultados de la búsqueda

Mostrando 1 - 10 de 16
  • Publicación
    Robust digital control for autonomous skid-steered agricultural robots
    (Elsevier, 2018-10) Fernández, Benjamín; Herrera Caro, Pedro Javier; Cerrada Somolinos, José Antonio
    There are two main issues to consider when designing a controller for autonomous off-road vehicles: velocity and terrain irregularities. Whereas the first one is measurable, the second one is very difficult to determine. Solutions to cover these issues could be very complex and difficult to implement in an embedded system with limited resources. The results obtained in our previous research for an adaptive approach implemented in a tractor with varying hitch forces, lead to the improvements presented here. This paper proposes a robust digital RST pole placement controller design for the lateral position, with sensitivity functions tuned to cover uncertainties and non-linearities not considered in the model. Simulations were implemented to assess the performance of the system and the controller implementation was applied to a skid-steered agricultural robot with limited computational resources and a state-of-the-art navigation system, which delivered satisfactory results.
  • Publicación
    A Simplified Optimal Path Following Controller for an Agricultural Skid-Steering Robot
    (Institute of Electrical and Electronics Engineers, 2019-08-02) Fernandez, Benjamin; Herrera Caro, Pedro Javier; Cerrada Somolinos, José Antonio
    The dynamics of a skid-steering robot present intrinsic non-linearities that make the design and implementation of a controller a very complex task, time-consuming, and difficult to implement into an embedded system with limited resources. This paper presents a simplified first order digital model approximation and an optimal observer-based control approach for the tracking of the lateral position of such robots. In order to verify the validity of this proposal, 3D real-time interactive simulations and real validations with an agricultural skid-steering robot were performed with satisfactory results.
  • Publicación
    A vision-based strategy to segment and localize ancient symbols written in stone
    (Springer, 2017-12-21) Duque Domingo, Jaime; Herrera Caro, Pedro Javier; Cerrada Somolinos, Carlos; Cerrada Somolinos, José Antonio; https://orcid.org/0000-0001-6649-5550; https://orcid.org/0000-0001-8679-6617; https://orcid.org/0000-0002-8591-6581; https://orcid.org/0000-0001-5492-5293
    This work proposes an automatic method to detect ancient symbols written in stone. The proposed method takes into account well-known techniques used in computer vision to identify the contour of the symbols in the image. The two-stage method consists of segmentation and localization processes. Segmentation process includes a pre-processing step, edge detection and thresholding. Localization process is based on two conditions that take into account several parameters, like the distance between points, and the orientation and the continuity of the edges. This proposal has been applied to localize Egyptian cartouches (borders enclosing the name of a king) and stonemason’s marks from images obtained under varying lighting conditions (controlled and natural lighting). The proposed method is compared favorably against other methods based on chain coding, neural networks and statistical correlation. The promising results give new possibilities to identify and recognize complex symbols and ancient texts.
  • Publicación
    Self-tuning Regulator for a Tractor with Varying Speed and Hitch Forces
    (Elsevier, 2018-02) Fernandez, Benjamin; Herrera Caro, Pedro Javier; Cerrada Somolinos, José Antonio
    Due to the changes in the soil, speed and hitch forces, the dynamics of a farm tractor are constantly changing making the design of an autonomous lane-tracking controller a very complex task. To be able to react to those changes, this paper presents a new adaptive system based on a self-tuning regulator made up of a recursive least-squares parameter identification algorithm for the plant combined with a minimum-degree pole placement (MDPP) method for changing the parameters of a digital RST controller in real time. The MDPP is computed by solving the Diophantine equation for the desired closed-loop reference model. The results presented show how the system is able to adapt the control parameters for different speeds and changes in the hitch cornering stiffness. As future work, this method could also be applied and assessed as a general controller, covering different sizes and different types of steering systems for off-road vehicles.
  • Publicación
    A Combined Model Based on Recurrent Neural Networks and Graph Convolutional Networks for Financial Time Series Forecasting
    (MDPI, 2023-01-02) Lazcano, Ana; Herrera Caro, Pedro Javier; Monge Moreno, Manuel Ángel
    Accurate and real-time forecasting of the price of oil plays an important role in the world economy. Research interest in forecasting this type of time series has increased considerably in recent decades, since, due to the characteristics of the time series, it was a complicated task with inaccurate results. Concretely, deep learning models such as Convolutional Neural Networks (CNNs) and Recurrent Neural Networks (RNNs) have appeared in this field with promising results compared to traditional approaches. To improve the performance of existing networks in time series forecasting, in this work two types of neural networks are brought together, combining the characteristics of a Graph Convolutional Network (GCN) and a Bidirectional Long Short-Term Memory (BiLSTM) network. This is a novel evolution that improves existing results in the literature and provides new possibilities in the analysis of time series. The results confirm a better performance of the combined BiLSTM-GCN approach compared to the BiLSTM and GCN models separately, as well as to the traditional models, with a lower error in all the error metrics used: the Root Mean Squared Error (RMSE), the Mean Squared Error (MSE), the Mean Absolute Percentage Error (MAPE) and the R-squared (R2). These results represent a smaller difference between the result returned by the model and the real value and, therefore, a greater precision in the predictions of this model.
  • Publicación
    On the Supervision of Peer Assessment Tasks: An Efficient Instructor Guidance Technique
    (Institute of Electrical and Electronics Engineers (IEEE), 2023-12) Hernández González, Jerónimo; Herrera Caro, Pedro Javier
    In peer assessment, students assess a task done by their peers, provide feedback and usually a grade. The extent to which these peer grades can be used to formally grade the task is unclear, with doubts often arising regarding their validity. The instructor could supervise the peer assessments, but would not then benefit from workload reduction, one of the most appealing features of peer assessment for instructors. Our proposal uses a probabilistic model to estimate a grade for each test, accounting for the degree of precision and bias of grading peers. The grade that the instructor would assign to a test can help enhance the model. Our main hypothesis is that guiding the instructor through supervision of a peer-assessed task by pointing out to them which test to evaluate next can lead to improvement in the validity of the model-estimated grades at an early stage. Moreover, the instructor can decide how many tests to grade based on their own criteria of tolerable uncertainty, as measured by the model. We validate the method using both synthetically generated data and real data collected in an actual class. Models that link the roles of the student as grading peer and as test-taker appear to better exploit available information, although simpler models are more appropriate in specific conditions. The best performing technique for guiding the instructor is that which selects the test with the highest expected entropy reduction. In general, empirical results are in line with the hypothesis of this study.
  • Publicación
    Deciphering Egyptian Hieroglyphs: Towards a New Strategy for Navigation in Museums
    (MDPI, 2017-04-14) Duque Domingo, Jaime; Herrera Caro, Pedro Javier; Valero Rodríguez, Enrique; Cerrada Somolinos, Carlos; Ministerio de Economía y Competitividad (España); Comunidad de Madrid
    This work presents a novel strategy to decipher fragments of Egyptian cartouches identifying the hieroglyphs of which they are composed. A cartouche is a drawing, usually inside an oval, that encloses a group of hieroglyphs representing the name of a monarch. Aiming to identify these drawings, the proposed method is based on several techniques frequently used in computer vision and consists of three main stages: first, a picture of the cartouche is taken as input and its contour is localized. In the second stage, each hieroglyph is individually extracted and identified. Finally, the cartouche is interpreted: the sequence of the hieroglyphs is established according to a previously generated benchmark. This sequence corresponds to the name of the king. Although this method was initially conceived to deal with both high and low relief writing in stone, it can be also applied to painted hieroglyphs. This approach is not affected by variable lighting conditions, or the intensity and the completeness of the objects. This proposal has been tested on images obtained from the Abydos King List and other Egyptian monuments and archaeological excavations. The promising results give new possibilities to recognize hieroglyphs, opening a new way to decipher longer texts and inscriptions, being particularly useful in museums and Egyptian environments. Additionally, devices used for acquiring visual information from cartouches (i.e., smartphones), can be part of a navigation system for museums where users are located in indoor environments by means of the combination of WiFi Positioning Systems (WPS) and depth cameras, as unveiled at the end of the document.
  • Publicación
    Reconocimiento facial en informativos televisivos mediante redes convolucionales profundas
    (Universidade da Coruña, 2025-09-01) Asensi-González, Ricardo; Herrera Caro, Pedro Javier
    Este trabajo propone un sistema de inteligencia artificial basado en redes neuronales profundas que permite la detección y reconocimiento de personas concretas en imágenes extraídas de informativos televisivos. Para ello, se ha creado un conjunto de datos (dataset) que consta de 12800 imágenes, centrado principalmente en figuras políticas de ámbito nacional. El sistema propuesto realiza la detección del individuo en la escena de manera automática utilizando la red YOLOv8 y, posteriormente, realiza su reconocimiento a partir del clasificador que proporcione mayor certidumbre. Para ello, se compararon siete arquitecturas de red neuronal convenientemente adaptadas a esta problemática concreta: VGG-16, VGG-19, InceptionV3, Xception, ResNet-101, MobileNetV2 y DenseNet-169, siendo este último el modelo que obtiene en promedio un mejor desempeño en todas las pruebas realizadas. Los resultados confirman la viabilidad del sistema y permiten sentar las bases para futuras investigaciones.
  • Publicación
    Detección y reconocimiento automático de jeroglíficos egipcios mediante visión por computador y validación cruzada
    (Comité Español de Automática, 2025-05-14) Herrera Caro, Pedro Javier; Duque-Domingo, Jaime; Fuentes-Ferrer, Raúl
    Los jeroglíficos egipcios forman parte de un lenguaje complejo, compuesto por cientos de símbolos que pueden transcribirse a fonemas latinos. Aunque existen estudios centrados en el reconocimiento de jeroglíficos mediante técnicas de visión por computador, este trabajo destaca por utilizar el mayor conjunto de datos de jeroglíficos, y que abarca los principales símbolos utilizados en estelas egipcias. Tras un análisis de los principales métodos de detección, segmentación y clasificación, se propone un enfoque combinado que, mediante el uso de técnicas clásicas de visión artificial para la segmentación basadas en regiones, y de votación mediante validación cruzada apoyada en una red ConvNeXt para la clasificación, ha sido evaluado en varias estelas de piedra talladas o pintadas con resultados prometedores. Este trabajo abre la puesta al desarrollo de técnicas automatizadas para la lectura de textos jeroglíficos egipcios.
  • Publicación
    Selección de candidatos usando lógica borrosa, PLN y aprendizaje profundo
    (Universidade da Coruña, 2025-09-01) González-González, Carlos; Herrera Caro, Pedro Javier
    La evaluación y selección de currículums es un proceso complejo en el que intervienen múltiples factores. Este trabajo presenta un modelo que integra dos componentes claramente diferenciados. En primer lugar, un sistema de evaluación multicriterio y selección difusa, basado en pesos que se autocalculan en función de diversas características. En segundo lugar, un sistema de procesamiento del lenguaje natural (PLN) junto con una red neuronal convolucional (CNN), que toma como entrada los resultados del primer sistema y clasifica los currículums en función de cinco criterios. La combinación de ambos enfoques permite reducir la subjetividad en el proceso de toma de decisiones durante la contratación, mejorando así la selección del candidato más adecuado para el puesto. Los resultados obtenidos se compararon con los derivados del proceso de análisis jerárquico (AHP) y con las decisiones de un grupo de expertos en recursos humanos. Para este estudio se utilizó una muestra de más de 1000 currículums