A Patch-Based Architecture for Multi-Label Classification from Single Positive Annotations

Warren Jouanneau, Warren Jouanneau, Aurélie Bugeau, Aurélie Bugeau, Marc Palyart, Nicolas Papadakis, Laurent Vézard

2023

Abstract

Supervised methods rely on correctly curated and annotated datasets. However, data annotation can be a cumbersome step needing costly hand labeling. In this paper, we tackle multi-label classification problems where only a single positive label is available in images of the dataset. This weakly supervised setting aims at simplifying datasets assembly by collecting only positive image exemples for each label without further annotation refinement. Our contributions are twofold. First, we introduce a light patch architecture based on the attention mechanism. Next, leveraging on patch embedding self-similarities, we provide a novel strategy for estimating negative examples and deal with positive and unlabeled learning problems. Experiments demonstrate that our architecture can be trained from scratch, whereas pre-training on similar databases is required for related methods from the literature.

Paper Citation

in Harvard Style

Jouanneau W., Bugeau A., Palyart M., Papadakis N. and Vézard L. (2023). A Patch-Based Architecture for Multi-Label Classification from Single Positive Annotations. In Proceedings of the 18th International Joint Conference on Computer Vision, Imaging and Computer Graphics Theory and Applications (VISIGRAPP 2023) - Volume 5: VISAPP; ISBN 978-989-758-634-7, SciTePress, pages 47-58. DOI: 10.5220/0011610800003417

in Bibtex Style

@conference{visapp23,
author={Warren Jouanneau and Aurélie Bugeau and Marc Palyart and Nicolas Papadakis and Laurent Vézard},
title={A Patch-Based Architecture for Multi-Label Classification from Single Positive Annotations},
booktitle={Proceedings of the 18th International Joint Conference on Computer Vision, Imaging and Computer Graphics Theory and Applications (VISIGRAPP 2023) - Volume 5: VISAPP},
year={2023},
pages={47-58},
publisher={SciTePress},
organization={INSTICC},
doi={10.5220/0011610800003417},
isbn={978-989-758-634-7},
}

in EndNote Style

TY - CONF

JO - Proceedings of the 18th International Joint Conference on Computer Vision, Imaging and Computer Graphics Theory and Applications (VISIGRAPP 2023) - Volume 5: VISAPP
TI - A Patch-Based Architecture for Multi-Label Classification from Single Positive Annotations
SN - 978-989-758-634-7
AU - Jouanneau W.
AU - Bugeau A.
AU - Palyart M.
AU - Papadakis N.
AU - Vézard L.
PY - 2023
SP - 47
EP - 58
DO - 10.5220/0011610800003417
PB - SciTePress