Information Fusion

Modification of information reduction processes in Convolutional Neural Networks

This work summarizes the main contributions developed during my PhD dissertation, devoted to the modification of information fusion processes in Convolutional Neural Networks.

Iosu Rodríguez Martínez, Humberto Bustince, Francisco Herrera, Zdenko Takáč

Funciones de pooling basadas en Penalty aplicadas a la reducción de características en Redes Neuronales Convolucionales

Presenting the extension of CombPool layers based on a “soft” version of Penalty functions on the XXI Congreso Español sobre Tecnologías y Lógica Fuzzy

Sep 4, 2022 8:00 PM — 9:00 PM Toledo, Spain

Replacing pooling functions in Convolutional Neural Networks by linear combinations of increasing functions

In this work, we replace traditional pooling by several alternative functions. In particular, we consider linear combinations of order statistics and generalizations of the Sugeno integral, extending the latter’s domain to the whole real line and setting the theoretical base for their application. We present an alternative pooling layer based on this strategy which we name “CombPool” layer.

Iosu Rodríguez Martínez, Julio Lafuente, Santiago Regivan, Graçaliz Dimuro, Francisco Herrero, Humberto Bustince