On Label Granularity and Object Localization

Abstract

Weakly supervised object localization (WSOL) aims to learn representations that encode object location using only image-level category labels. However, many objects can be labeled at different levels of granularity. Is it an animal, a bird, or a great horned owl? Which image-level labels should we use? In this paper we study the role of label granularity in WSOL. To facilitate this investigation we introduce iNatLoc500, a new large-scale fine-grained benchmark dataset for WSOL. Surprisingly, we find that choosing the right training label granularity provides a much larger performance boost than choosing the best WSOL algorithm. We also show that changing the label granularity can significantly improve data efficiency.

Cite

Text

Cole et al. "On Label Granularity and Object Localization." Proceedings of the European Conference on Computer Vision (ECCV), 2022. doi:10.1007/978-3-031-20080-9_35

Markdown

[Cole et al. "On Label Granularity and Object Localization." Proceedings of the European Conference on Computer Vision (ECCV), 2022.](https://mlanthology.org/eccv/2022/cole2022eccv-label/) doi:10.1007/978-3-031-20080-9_35

BibTeX

@inproceedings{cole2022eccv-label,
  title     = {{On Label Granularity and Object Localization}},
  author    = {Cole, Elijah and Wilber, Kimberly and Van Horn, Grant and Yang, Xuan and Fornoni, Marco and Perona, Pietro and Belongie, Serge and Howard, Andrew and Aodha, Oisin Mac},
  booktitle = {Proceedings of the European Conference on Computer Vision (ECCV)},
  year      = {2022},
  doi       = {10.1007/978-3-031-20080-9_35},
  url       = {https://mlanthology.org/eccv/2022/cole2022eccv-label/}
}