Examinando por Autor "Plaza, Javier"

Mostrando 1 - 3 de 3
  • Miniatura
    Publicación
    Distributed Deep Learning for Remote Sensing Data Interpretation
    (IEEE, 2021-03-15) Haut, Juan Mario; Paoletti, Mercedes Eugenia; Moreno Álvarez, Sergio; Plaza, Javier; Rico Gallego, Juan Antonio; Plaza, Antonio; https://orcid.org/0000-0001-6701-961X; https://orcid.org/0000-0003-1030-3729; https://orcid.org/0000-0002-2384-9141; https://orcid.org/0000-0002-4264-7473; https://orcid.org/0000-0002-9613-1659
    As a newly emerging technology, deep learning (DL) is a very promising field in big data applications. Remote sensing often involves huge data volumes obtained daily by numerous in-orbit satellites. This makes it a perfect target area for data-driven applications. Nowadays, technological advances in terms of software and hardware have a noticeable impact on Earth observation applications, more specifically in remote sensing techniques and procedures, allowing for the acquisition of data sets with greater quality at higher acquisition ratios. This results in the collection of huge amounts of remotely sensed data, characterized by their large spatial resolution (in terms of the number of pixels per scene), and very high spectral dimensionality, with hundreds or even thousands of spectral bands. As a result, remote sensing instruments on spaceborne and airborne platforms are now generating data cubes with extremely high dimensionality, imposing several restrictions in terms of both processing runtimes and storage capacity. In this article, we provide a comprehensive review of the state of the art in DL for remote sensing data interpretation, analyzing the strengths and weaknesses of the most widely used techniques in the literature, as well as an exhaustive description of their parallel and distributed implementations (with a particular focus on those conducted using cloud computing systems). We also provide quantitative results, offering an assessment of a DL technique in a specific case study (source code available: https://github.com/mhaut/cloud-dnn-HSI). This article concludes with some remarks and hints about future challenges in the application of DL techniques to distributed remote sensing data interpretation problems. We emphasize the role of the cloud in providing a powerful architecture that is now able to manage vast amounts of remotely sensed data due to its implementation simplicity, low cost, and high efficiency compared to other parallel and distributed architectures, such as grid computing or dedicated clusters.
  • Miniatura
    Publicación
    Evaluación de Rendimiento del Entrenamiento Distribuido de Redes Neuronales Profundas en Plataformas Heterogéneas
    (Universidad de Extremadura, 2019) Moreno Álvarez, Sergio; Paoletti, Mercedes Eugenia; Haut, Juan Mario; Rico Gallego, Juan Antonio; Plaza, Javier; Díaz Martín, Juan Carlos; Vega Rodriguez, Miguel ángel; Plaza Miguel, Antonio J.; https://orcid.org/0000-0003-1030-3729; https://orcid.org/0000-0001-6701-961X; https://orcid.org/0000-0002-4264-7473; https://orcid.org/0000-0002-8908-1606; https://orcid.org/0000-0002-8435-3844
    Asynchronous stochastic gradient descent es una tecnica de optimizacion comunmente utilizada en el entrenamiento distribuido de redes neuronales profundas. En distribuciones basadas en particionamiento de datos, se entrena una replica del modelo en cada unidad de procesamiento de la plataforma, utilizando conjuntos de muestras denominados mini-batches. Este es un proceso iterativo en el que al nal de cada mini-batch, las replicas combinan los gradientes calculados para actualizar su copia local de los parametros. Sin embargo, al utilizar asincronismo, las diferencias en el tiempo de entrenamiento por iteracion entre replicas provocan la aparicion del staleness, esto es, las replicas progresan a diferente velocidad y en el entrenamiento de cada replica se utiliza una vers on no actualizada de los parametros. Un alto gradde staleness tiene un impacto negativo en la precision del modelo resultante. Ademas, las plataformas de computacion de alto rendimiento suelen ser heterogeneas, compuestas por CPUs y GPUs de diferentes capacidades, lo que agrava el problema de staleness. En este trabajo, se propone aplicar t ecnicas de equilibrio de carga computacional, bien conocidas en el campo de la Computaci on de Altas Prestaciones, al entrenamiento distribuido de modelos profundos. A cada r eplica se asignar a un n umero de mini-batches en proporci on a su velocidad relativa. Los resultados experimentales obtenidos en una plataforma hete-rog enea muestran que, si bien la precisi on se mantiene constante, el rendimiento del entrenamiento aumenta considerablemente, o desde otro punto de vista, en el mismo tiempo de entrenamiento, se alcanza una mayor precisi on en las estimaciones del modelo. Discutimos las causas de tal incremento en el rendimiento y proponemos los pr oximos pasos para futuras investigaciones.
  • Miniatura
    Publicación
    Training deep neural networks: a static load balancing approach
    (Springer, 2020-03-02) Moreno Álvarez, Sergio; Haut, Juan Mario; Paoletti, Mercedes Eugenia; Rico Gallego, Juan Antonio; Díaz Martín, Juan Carlos; Plaza, Javier; https://orcid.org/0000-0003-1030-3729; https://orcid.org/0000-0002-4264-7473; https://orcid.org/0000-0002-8435-3844; https://orcid.org/0000-0002-8908-1606
    Deep neural networks are currently trained under data-parallel setups on high-performance computing (HPC) platforms, so that a replica of the full model is charged to each computational resource using non-overlapped subsets known as batches. Replicas combine the computed gradients to update their local copies at the end of each batch. However, differences in performance of resources assigned to replicas in current heterogeneous platforms induce waiting times when synchronously combining gradients, leading to an overall performance degradation. Albeit asynchronous communication of gradients has been proposed as an alternative, it suffers from the so-called staleness problem. This is due to the fact that the training in each replica is computed using a stale version of the parameters, which negatively impacts the accuracy of the resulting model. In this work, we study the application of well-known HPC static load balancing techniques to the distributed training of deep models. Our approach is assigning a different batch size to each replica, proportional to its relative computing capacity, hence minimizing the staleness problem. Our experimental results (obtained in the context of a remotely sensed hyperspectral image processing application) show that, while the classification accuracy is kept constant, the training time substantially decreases with respect to unbalanced training. This is illustrated using heterogeneous computing platforms, made up of CPUs and GPUs with different performance.