skip to main content
Invitado
Mi portal
Mi Cuenta
Cerrar sesión
Identificarse
This feature requires javascript
Tags
Periódicos Eletrónicos
Libros Eletrónicos
Bases de Datos
Bibliotecas de USP
Ayuda
Ayuda
Idioma:
Inglés
Castellano
Portugués (Brasil)
This feature required javascript
This feature requires javascript
Primo Search
Búsqueda General
Búsqueda General
Colección Física
Colecciones Físicas
Producción Intelectual USP
Producción USP
Search For:
Clear Search Box
Search in:
Búsqueda General
Or select another collection:
Search in:
Búsqueda General
Búsqueda Avanzada
Búsqueda por Índices
This feature requires javascript
Tipo de recurso
criteria input
en cualquier parte del registro
en el título
como autor/creador
en materias
Fecha de creación
lsr01
lsr02
lsr03
lsr04
Orientador
Show Results with:
en el título
Show Results with:
en cualquier parte del registro
en el título
como autor/creador
en materias
Fecha de creación
lsr01
lsr02
lsr03
lsr04
Orientador
Ver resultados con:
criteria input
que contiene las palabras de la búsqueda
con la frase exacta
empieza por
Ver resultados con:
Sumario
criteria input
Y
O
NO
This feature requires javascript
Features transfer learning between domains for image and video recognition tasks
Santos, Fernando Pereira Dos
Biblioteca Digital de Teses e Dissertações da USP; Universidade de São Paulo; Instituto de Ciências Matemáticas e de Computação 2020-01-24
Acceso en línea
Recurso en línea
Detalles
Comentarios y Etiquetas
Servicios adicionales
This feature requires javascript
Acciones
Agregar a Mi Portal
Eliminar de Mi Portal
Correo Electrónico
Imprimir
Enlae permanente
Cita bibliográfica
EasyBib
EndNote
RefWorks
Delicious
Exportación RIS
Exportar BibTeX
This feature requires javascript
Título:
Features transfer learning between domains for image and video recognition tasks
Autor:
Santos, Fernando Pereira Dos
Orientador:
Ponti, Moacir Antonelli
Materias:
Alinhamento De Variedades
;
Medidas De Generalização
;
Transferência De Aprendizado De Características
;
Cruzamento De Domínios
;
Aprendizado Profundo
;
Deep Learning
;
Cross-Domain
;
Generalization Measures
;
Manifold Alignment
;
Features Transfer Learning
Notas:
Tese (Doutorado)
Descripción:
Feature transfer learning aims to reuse knowledge previously acquired in some source dataset to apply it in another target data and/or task. A requirement for the transfer of knowledge is the quality of feature spaces obtained, in which deep learning methods are widely applied since those provide discriminative and general descriptors. In this context, the main questions include: what to transfer align the data distribution from source and target, and adjusting the parameters to increase the models generalization capability; how to transfer investigating methods that work on the features spaces or also on the learned models; and when to transfer studying which datasets are mode adequate for transferring, considering discrepancies between source and target data, such as they different acquisition settings, clutter and illumination variation, among others. This thesis advocates that the focus should be in transferring feature spaces, learned by convolutional neural networks, in particular investigating the descriptive potential of inner and initial layers of such deep convolutional networks, and the approximation of feature spaces before aligning the data distribution in order to allow for better solutions, as well as the use of both labeled and unlabeled for feature learning. Besides the transfer learning methods, such as fine-tuning and manifold alignment, with use of classical evaluation metrics for recognition performance, a generalization metric between domains is also proposed to evaluate transfer learning. This thesis contributes with: an analysis of multiple descriptors contained in supervised deep networks; a new architecture with a loss function for semi-supervised deep networks (Weighted Label Loss), in which all available data, labeled or unlabeled, are incorporated to provide learning; and a new generalization metric (Cross-domain Feature Space Generalization Measure) that can be applied to any model and evaluation system
DOI:
10.11606/T.55.2020.tde-19032020-083537
Editor:
Biblioteca Digital de Teses e Dissertações da USP; Universidade de São Paulo; Instituto de Ciências Matemáticas e de Computação
Fecha de creación:
2020-01-24
Formato:
Adobe PDF
Idioma:
Inglés
Enlaces
Este registro en el Catálogo
Teses e Dissertações USP
Acesso ao doi
E-mail do autor
E-mail do orientador
This feature requires javascript
This feature requires javascript
Volver a la lista de resultados
Anterior
Resultado
3
Siguiente
This feature requires javascript
This feature requires javascript
Buscando en bases de datos remotas, por favor espere
Buscando por
en
scope:(USP_PRODUCAO),scope:(USP_EBOOKS),scope:("PRIMO"),scope:(USP),scope:(USP_EREVISTAS),scope:(USP_FISICO),primo_central_multiple_fe
Mostrar lo que tiene hasta ahora
This feature requires javascript
This feature requires javascript