Examinando por Autor "Tellaeche Iglesias, Alberto"

Mostrando 1 - 7 de 7
  • Miniatura
    Ítem
    Comparative benchmark of sampling-based and DRL motion planning methods for industrial robotic arms
    (Multidisciplinary Digital Publishing Institute (MDPI), 2025-08-25) Fidalgo Astorquia, Ignacio; Villate Castillo, Guillermo; Tellaeche Iglesias, Alberto; Vazquez, Juan-Ignacio
    This study presents a comprehensive comparison between classical sampling-based motion planners from the Open Motion Planning Library (OMPL) and a learning-based planner based on Soft Actor–Critic (SAC) for motion planning in industrial robotic arms. Using a UR3e robot equipped with an RG2 gripper, we constructed a large-scale dataset of over 100,000 collision-free trajectories generated with MoveIt-integrated OMPL planners. These trajectories were used to train a DRL agent via curriculum learning and expert demonstrations. Both approaches were evaluated on key metrics such as planning time, success rate, and trajectory smoothness. Results show that the DRL-based planner achieves higher success rates and significantly lower planning times, producing more compact and deterministic trajectories. Time-optimal parameterization using TOPPRA ensured the dynamic feasibility of all trajectories. While classical planners retain advantages in zero-shot adaptability and environmental generality, our findings highlight the potential of DRL for real-time and high-throughput motion planning in industrial contexts. This work provides practical insights into the trade-offs between traditional and learning-based planning paradigms, paving the way for hybrid architectures that combine their strengths.
  • No hay miniatura disponible
    Ítem
    Deep learning applications on cybersecurity: a practical approach
    (Elsevier B.V., 2024-01-01) Miranda García, Alberto; Zubillaga Rego, Agustín José; Pastor López, Iker; Sanz Urquijo, Borja; Tellaeche Iglesias, Alberto; Gaviria de la Puerta, José; García Bringas, Pablo
    One of the most difficult challenges for computer systems has been security. On the other hand, new developments in machine learning are having an impact on almost every aspect of computer science, including cybersecurity. To analyze this impact, we have created three distinct cybersecurity-related problems to show the advantages of deep learning techniques. We examined their potential applications for SPAM filtering, detecting malicious software, and adult-content detection. We experimented with various techniques, such as Long Short-Term Memory (LSTMs) for spam filtering, Deep Neural Networks (DNNs) for malware detection, Convolutional Neural Networks (CNNs) combined with Transfer Learning for adult content detection and image augmentation methods. We are able to achieve an Area Under ROC Curve greater than 0.94 in every scenario, proving that excellent performance with a good relation between cost and effectiveness may be created without the need of complex designs.
  • Miniatura
    Ítem
    Gesture-based human machine interaction using RCNNs in limited computation power devices
    (NLM (Medline), 2021-12-08) Tellaeche Iglesias, Alberto ; Fidalgo Astorquia, Ignacio ; Vazquez, Juan-Ignacio ; Saikia, Surajit
    The use of gestures is one of the main forms of human machine interaction (HMI) in many fields, from advanced robotics industrial setups, to multimedia devices at home. Almost every gesture detection system uses computer vision as the fundamental technology, with the already well-known problems of image processing: changes in lighting conditions, partial occlusions, variations in color, among others. To solve all these potential issues, deep learning techniques have been proven to be very effective. This research proposes a hand gesture recognition system based on convolutional neural networks and color images that is robust against environmental variations, has a real time performance in embedded systems, and solves the principal problems presented in the previous paragraph. A new CNN network has been specifically designed with a small architecture in terms of number of layers and total number of neurons to be used in computationally limited devices. The obtained results achieve a percentage of success of 96.92% on average, a better score than those obtained by previous algorithms discussed in the state of the art.
  • No hay miniatura disponible
    Ítem
    An innovative framework for supporting content-based authorship identification and analysis in social media networks
    (Oxford University Press, 2024-08) Gaviria de la Puerta, José; Pastor López, Iker; Tellaeche Iglesias, Alberto; Sanz Urquijo, Borja; Sanjurjo González, Hugo; Cuzzocrea, Alfredo; Bringas García, Pablo
    Content-based authorship identification is an emerging research problem in online social media networks, due to a wide collection of issues ranging from security to privacy preservation, from radicalization to defamation detection, and so forth. Indeed, this research has attracted a relevant amount of attention from the research community during the past years. The general problem becomes harder when we consider the additional constraint of identifying the same false profile over different social media networks, under obvious considerations. Inspired by this emerging research challenge, in this paper we propose and experimentally assess an innovative framework for supporting content-based authorship identification and analysis in social media networks.
  • Miniatura
    Ítem
    On combining convolutional autoencoders and support vector machines for fault detection in industrial textures
    (MDPI AG, 2021-05-02) Tellaeche Iglesias, Alberto ; Campos Anaya, Miguel Ángel; Pajares Martinsanz, Gonzalo; Pastor López, Iker
    Defects in textured materials present a great variability, usually requiring ad‐hoc solutions for each specific case. This research work proposes a solution that combines two machine learning-based approaches, convolutional autoencoders, CA; one class support vector machines, SVM. Both methods are trained using only defect free textured images for each type of analyzed texture, label-ing the samples for the SVMs in an automatic way. This work is based on two image processing streams using image sensors: (1) the CA first processes the incoming image from the input to the output, producing a reconstructed image, from which a measurement of correct or defective image is obtained; (2) the second process uses the latent layer information as input to the SVM to produce a measurement of classification. Both measurements are effectively combined, making an additional research contribution. The results obtained achieve a percentage of success of 92% on average, out-performing results of previous works.
  • Miniatura
    Ítem
    A practical approach on performance assessment of federated learning algorithms for defect detection in industrial applications
    (Institute of Electrical and Electronics Engineers Inc., 2023-09-29) Zuluaga, Eduard; Jaziri, Sondos; Tellaeche Iglesias, Alberto ; Pastor López, Iker
    One of the most common problems to be solved in industrial environments is the detection of defects in the manufacturing quality control of different products. The design of such automatic systems, especially if they are based on image processing, often presents difficulties related to the availability of sufficient well- labeled data for initial training. Usually, it is not easy to have enough industrial samples with defects to have a sufficiently large and balanced dataset. This research work presents a new method of hybridisation of convolutional neural networks used for defect detection in industry by means of image processing, using Federated Learning (FL) techniques. This method is able to overcome the limitations and problems presented by this type of systems stated above. In this research, it is demonstrated with examples of defect detection in textures that it is possible to reach a defect detection effectiveness above 90% on average, in problems where no dataset is available, using federated learning algorithms with classifiers based on Convolutional Neural Networks previously trained in other problems of defect detection in other types of textures. The creation of such systems for error detection on previously untrained data using federated learning and obtaining this effectiveness represents an advance on the state of the art with respect to the existing approaches, and constitutes the main contribution of this research work.
  • No hay miniatura disponible
    Ítem
    Quality assessment methodology based on machine learning with small datasets: industrial castings defects
    (Elsevier B.V., 2021-10-07) Pastor López, Iker; Sanz Urquijo, Borja; Tellaeche Iglesias, Alberto; Psaila, Giuseppe; Gaviria de la Puerta, José; García Bringas, Pablo
    Nowadays there are numerous problems for which use of a multi-objective in image classification would be desirable although, unfortunately, the number of samples is too low. In these situations, higher level classifications could also work (for example, in surface defect detection, it is important to identify the defect, but it could also be useful to detect whether or not the object has a defect). To this end, we present a methodology called BoDoC which allows to improve this classification. To evaluate the methodology, we have created a new dataset, obtained from a foundry, to detect surface errors in casting pieces with 2 different defects: (i) inclusions, (ii) coldlaps and (iii) misruns. We also present a collection of techniques to select featu res from the images. We prove that our methodology improves the direct classification results in real world scenarios, with 91.305% precision.