Caracterización y análisis de los modelos de reconocimiento de imágenes mediante redes neuronales convolucionales

Abstract

This research is based on the objective of studying the main algorithms or models that allow recognizing images from the use of convolutional neural networks, through a documentary review, delving into their possible uses and limitations. this through the study of notions such as Neural Networks, their architecture, application, convolutions, the mathematical models that support them and their field of application, the functioning of a neural network, among others; Following this, the characterization of the main existing models and algorithms focused on image recognition from convolutional networks will be addressed, with special emphasis on their technical component; thus laying the foundations for the third section, the identification of the possible uses of the models and algorithms characterized, emphasizing practical cases of their application and, finally, the analysis of the main challenges and limitations of the models and algorithms identified, emphasizing their technical aspect.