• Channel redundancy and overlap in convolutional neural networks with channel-wise NNK graphs 

      Bonet Solé, David; Ortega, Antonio; Ruiz Hidalgo, Javier; Shekkizhar, Sarath (Institute of Electrical and Electronics Engineers (IEEE), 2022)
      Text en actes de congrés
      Accés obert
      Feature spaces in the deep layers of convolutional neural networks (CNNs) are often very high-dimensional and difficult to inter-pret. However, convolutional layers consist of multiple channels that are activated by different ...
    • Channel-wise early stopping without a validation set via NNK polytope interpolation 

      Bonet Solé, David; Ortega, Antonio; Ruiz Hidalgo, Javier; Sarath Shekkizhar, Sarath (2021)
      Text en actes de congrés
      Accés obert
      State-of-the-art neural network architectures continue to scale in size and deliver impressive generalization results, although this comes at the expense of limited interpretability. In particular, a key challenge is to ...
    • Modeling of contours in wavelet domain for generalized lifting image compression 

      Rolon Garrido, Julio Cesar; Ortega, Antonio; Salembier Clairon, Philippe Jean (2009)
      Text en actes de congrés
      Accés obert
      This paper introduces the design of context-based models of contours in the wavelet domain, which are used to construct generalized lifting (GL) mappings for image compression. The GL context-based mapping may significantly ...
    • Performance evaluation of probability density estimators for unsupervised information theoretical region merging 

      Calderero Patino, Felipe; Marqués Acosta, Fernando; Ortega, Antonio (2009)
      Comunicació de congrés
      Accés obert
      Information theoretical region merging techniques have been shown to provide a state-of-the-art unified solution for natural and texture image segmentation. Here, we study how the segmentation results can be further ...
    • Study of manifold geometry using multiscale non-negative kernel graphs 

      Hurtado Gómez, Carlos; Shekkizhar, Sarath; Ruiz Hidalgo, Javier; Ortega, Antonio (Institute of Electrical and Electronics Engineers (IEEE), 2023)
      Text en actes de congrés
      Accés obert
      Modern machine learning systems are increasingly trained on large amounts of data embedded in high-dimensional spaces. Often this is done without analyzing the structure of the dataset. In this work, we propose a framework ...