Semi-supervised Segmentation Based on Error-Correcting Supervision

Registro completo de metadados
MetadadosDescriçãoIdioma
Autor(es): dc.contributorOstbayerische Technische Hochschule Regensburg-
Autor(es): dc.contributorUniversidade Federal de São Carlos (UFSCar)-
Autor(es): dc.contributorUniversidade Estadual Paulista (UNESP)-
Autor(es): dc.creatorMendel, Robert-
Autor(es): dc.creatorde Souza, Luis Antonio-
Autor(es): dc.creatorRauber, David-
Autor(es): dc.creatorPapa, João Paulo-
Autor(es): dc.creatorPalm, Christoph-
Data de aceite: dc.date.accessioned2025-08-21T21:04:53Z-
Data de disponibilização: dc.date.available2025-08-21T21:04:53Z-
Data de envio: dc.date.issued2022-05-01-
Data de envio: dc.date.issued2022-05-01-
Data de envio: dc.date.issued2019-12-31-
Fonte completa do material: dc.identifierhttp://dx.doi.org/10.1007/978-3-030-58526-6_9-
Fonte completa do material: dc.identifierhttp://hdl.handle.net/11449/233045-
Fonte: dc.identifier.urihttp://educapes.capes.gov.br/handle/11449/233045-
Descrição: dc.descriptionPixel-level classification is an essential part of computer vision. For learning from labeled data, many powerful deep learning models have been developed recently. In this work, we augment such supervised segmentation models by allowing them to learn from unlabeled data. Our semi-supervised approach, termed Error-Correcting Supervision, leverages a collaborative strategy. Apart from the supervised training on the labeled data, the segmentation network is judged by an additional network. The secondary correction network learns on the labeled data to optimally spot correct predictions, as well as to amend incorrect ones. As auxiliary regularization term, the corrector directly influences the supervised training of the segmentation network. On unlabeled data, the output of the correction network is essential to create a proxy for the unknown truth. The corrector’s output is combined with the segmentation network’s prediction to form the new target. We propose a loss function that incorporates both the pseudo-labels as well as the predictive certainty of the correction network. Our approach can easily be added to supervised segmentation models. We show consistent improvements over a supervised baseline on experiments on both the Pascal VOC 2012 and the Cityscapes datasets with varying amounts of labeled data.-
Descrição: dc.descriptionOstbayerische Technische Hochschule Regensburg-
Descrição: dc.descriptionFederal University of São Carlos-
Descrição: dc.descriptionSão Paulo State University-
Descrição: dc.descriptionSão Paulo State University-
Formato: dc.format141-157-
Idioma: dc.languageen-
Relação: dc.relationLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)-
???dc.source???: dc.sourceScopus-
Título: dc.titleSemi-supervised Segmentation Based on Error-Correcting Supervision-
Tipo de arquivo: dc.typeaula digital-
Aparece nas coleções:Repositório Institucional - Unesp

Não existem arquivos associados a este item.