CaltechAUTHORS
Published July 20, 2022 | Version public
Discussion Paper Metadata-only

On Label Granularity and Object Localization

Creators

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.

Additional Information

We thank the iNaturalist community for sharing images and species annotations. This work was supported by the Caltech Resnick Sustainability Institute, an NSF Graduate Research Fellowship (grant number DGE1745301), and the Pioneer Centre for AI (DNRF grant number P1).

Additional details

Identifiers

Eprint ID
118461
Resolver ID
CaltechAUTHORS:20221219-234038678

Related works

Describes
http://arxiv.org/abs/2207.10225 (URL)

Funding

Resnick Sustainability Institute
NSF Graduate Research Fellowship
DGE-1745301
Danish National Research Foundation
P1

Dates

Created
2022-12-21
Created from EPrint's datestamp field
Updated
2023-06-02
Created from EPrint's last_modified field

Caltech Custom Metadata

Caltech groups
Resnick Sustainability Institute