Unsupervised Discovery of the Long-Tail in Instance Segmentation Using Hierarchical Self-Supervision
Unsupervised Discovery of the Long-Tail in Instance Segmentation Using Hierarchical Self-Supervision
About this item
Full title
Author / Creator
Publisher
Ithaca: Cornell University Library, arXiv.org
Journal title
Language
English
Formats
Publication information
Publisher
Ithaca: Cornell University Library, arXiv.org
Subjects
More information
Scope and Contents
Contents
Instance segmentation is an active topic in computer vision that is usually solved by using supervised learning approaches over very large datasets composed of object level masks. Obtaining such a dataset for any new domain can be very expensive and time-consuming. In addition, models trained on certain annotated categories do not generalize well t...
Alternative Titles
Full title
Unsupervised Discovery of the Long-Tail in Instance Segmentation Using Hierarchical Self-Supervision
Authors, Artists and Contributors
Author / Creator
Identifiers
Primary Identifiers
Record Identifier
TN_cdi_proquest_journals_2509122089
Permalink
https://devfeature-collection.sl.nsw.gov.au/record/TN_cdi_proquest_journals_2509122089
Other Identifiers
E-ISSN
2331-8422
