Visual attention and novelty detection: experiments with automatic scale selection

Registro completo de metadados
MetadadosDescriçãoIdioma
Autor(es): dc.creatorNehmzow, Ulrich-
Autor(es): dc.creatorVieira Neto, Hugo-
Data de aceite: dc.date.accessioned2025-08-29T11:54:44Z-
Data de disponibilização: dc.date.available2025-08-29T11:54:44Z-
Data de envio: dc.date.issued2013-11-21-
Data de envio: dc.date.issued2006-
Fonte completa do material: dc.identifierhttp://repositorio.utfpr.edu.br/jspui/handle/1/660-
Fonte: dc.identifier.urihttp://educapes.capes.gov.br/handle/capes/1085552-
Descrição: dc.descriptionWe present experiments with an autonomous inspection robot, whose task was to highlight novel features in its environment using camera images. Experiments were conducted with two different attention mechanisms — saliency map and multiscale Harris detector — and two different novelty detection mechanisms — the Grow-When- Required neural network and incremental PCA. For both mechanisms we compared fixed-scale image encoding with automatically scaled image patches. Results show that using automatic scale selection provides a more efficient representation of the visual input space, but that performance is generally better using a fixed-scale image encoding.-
Descrição: dc.description5000-
Formato: dc.formatapplication/pdf-
Idioma: dc.languagept_BR-
Publicador: dc.publisherCuritiba-
Relação: dc.relationTowards Autonomous Robotic Systems-
Relação: dc.relationhttp://citeseerx.ist.psu.edu/viewdoc/download?doi=10.1.1.61.8433&rep=rep1&type=pdf-
Palavras-chave: dc.subjectRobôs móveis-
Palavras-chave: dc.subjectRedes neurais (Computação)-
Palavras-chave: dc.subjectVisão por computador-
Palavras-chave: dc.subjectMobile robots-
Palavras-chave: dc.subjectComputer vision-
Palavras-chave: dc.subjectNeural networks (Computer science)-
Título: dc.titleVisual attention and novelty detection: experiments with automatic scale selection-
Tipo de arquivo: dc.typeaula digital-
Aparece nas coleções:Repositorio Institucional da UTFPR - RIUT

Não existem arquivos associados a este item.