Please use this identifier to cite or link to this item: http://hdl.handle.net/10609/126709
Title: Detección y clasificación de células anormales de sangre periférica usando técnicas de N-Shot Learning
Author: Lucas Guerrero, José
Tutor: Alférez, Santiago  
Others: Prados Carrasco, Ferran  
Abstract: Using Machine Learning technologies for digital image processing, it is possible to develop applications able to classify images between different categories. In this master's thesis, Deep Learning models of the n-shot learning type have been implemented for the classification of peripheral blood images. These techniques are suitable when not enough sample is available for the training process. A Siamese Network has been developed for checking if two different images correspond to the same cell type. A five-shot network has been developed to classify leukocytes between healthy and affected by Burkkit's Lymphoma. Both models were developed in PyTorch and Fast.ai obtaining success rates higher than 75% and 85% respectively. The conclusion of the work is that few shot techniques are a good approach when not enough sample is available to apply conventional techniques.
Keywords: deep learning
n-shot learning
lymphoids cells
lymphomas
Document type: info:eu-repo/semantics/masterThesis
Issue Date: 2-Jan-2021
Publication license: http://creativecommons.org/licenses/by-nc-nd/3.0/es/  
Appears in Collections:Trabajos finales de carrera, trabajos de investigación, etc.

Files in This Item:
File Description SizeFormat 
pepelucasTFM0121memoria.pdfMemoria del TFM1,43 MBAdobe PDFThumbnail
View/Open