| Texto completo | |
| Autor(es): |
Nunes, Ian
;
Pereira, Matheus B.
;
Oliveira, Hugo
;
dos Santos, Jefersson A.
;
Poggi, Marcus
;
IEEE
Número total de Autores: 6
|
| Tipo de documento: | Artigo Científico |
| Fonte: | 2022 IEEE INTERNATIONAL CONFERENCE ON IMAGE PROCESSING, ICIP; v. N/A, p. 5-pg., 2022-01-01. |
| Resumo | |
Open set segmentation is a relatively new and unexplored task, with just a handful of methods proposed to model such tasks. We propose a novel method called CoReSeg that tackles the issue using class conditioned reconstruction of the input images according to their pixelwise mask. Our method conditions each input pixel to all known classes, expecting higher errors for pixels of unknown classes. It was observed that the proposed method produces better semantic consistency in its predictions than the baselines, resulting in cleaner segmentation maps that better fit object boundaries. CoReSeg outperforms state-of-the-art methods on the Vaihingen and Potsdam ISPRS datasets, while also being competitive on the Houston 2018 IEEE GRSS Data Fusion dataset. Our official implementation for CoReSeg is available at: https://github.com/iannunes/CoReSeg. (AU) | |
| Processo FAPESP: | 20/06744-5 - Deep learning e representações intermediárias para análise de imagens pediátricas |
| Beneficiário: | Hugo Neves de Oliveira |
| Modalidade de apoio: | Bolsas no Brasil - Pós-Doutorado |
| Processo FAPESP: | 15/22308-2 - Representações intermediárias em Ciência Computacional para descoberta de conhecimento |
| Beneficiário: | Roberto Marcondes Cesar Junior |
| Modalidade de apoio: | Auxílio à Pesquisa - Temático |