@InProceedings{AlvarengaeSilvaAlme:2023:OpSeDo,
author = "Alvarenga e Silva, Lucas Fernando and Almeida, Jurandy",
affiliation = "{Universidade Estadual de Campinas – UNICAMP} and {Universidade
Federal de S{\~a}o Carlos – UFScar}",
title = "Open Set Domain Adaptation Methods in Deep Networks for Image
Recognition",
booktitle = "Proceedings...",
year = "2023",
editor = "Clua, Esteban Walter Gonzalez and K{\"o}rting, Thales Sehn and
Paulovich, Fernando Vieira and Feris, Rogerio",
organization = "Conference on Graphics, Patterns and Images, 36. (SIBGRAPI)",
keywords = "open set domain adaptation, unsupervised domain adaptation, domain
adaptation, deep learning.",
abstract = "Deep learning (DL) has revolutionized various fields through its
remarkable capacity to learn from raw data. However, in
uncontrolled environments like in the wild, the performance of
these systems might degrade to some extent, especially with
unlabeled datasets. Naive approaches train DL models on labeled
datasets (source domains) that resemble the unlabeled test dataset
(target domain), but nonetheless, this approach may not yield
optimal results due to domain and category-shift problems. These
issues have been the primary focus of Unsupervised Domain
Adaptation (UDA) and Open Set Recognition research areas. To
address the domain-shift problem, we introduced the Multi-Source
Domain Alignment Layers (MS-DIAL), a structural solution for
multi-source UDA. MS-DIAL aligns the source domains and the target
domain at various levels of the feature space, individually
achieving competitive results comparable to the state-of-the-art,
and when combined with other UDA methods, it further enhances
transferability by up to 30.64% in relative performance gains.
Subsequently, we tackled the demanding setup of Open Set Domain
Adaptation (OSDA), where both domain and category-shift issues
coexist. Our proposed approach involves dealing with negatives,
extracting a high-confidence set of unknown instances, and using
them as a hard constraint to refine the classification boundaries
of OSDA methods. We assessed our proposal in an extensive set of
experiments, which achieved up to 5.8% of absolute performance
gains.",
conference-location = "Rio Grande, RS",
conference-year = "Nov. 06-09, 2023",
language = "en",
ibi = "8JMKD3MGPEW34M/49S978P",
url = "http://urlib.net/ibi/8JMKD3MGPEW34M/49S978P",
targetfile = "silva13.pdf",
urlaccessdate = "2024, May 05"
}