You're currently viewing an old version of this dataset. To see the current version, click here.

Probe-R

This dataset was created using Visual Genome (VG) [21]. The goal is to have high accuracy and confidence in selecting the correct prompt for the respective image where the incorrect prompt has either the relation or the object swapped.

Data and Resources

Cite this as

Madeline Chantry Schiappa, Michael Cogswell, Ajay Divakaran, Yogesh S. Rawat (2024). Dataset: Probe-R. https://doi.org/10.57702/5kx18ee7

DOI retrieved: December 2, 2024

Additional Info

Field Value
Created December 2, 2024
Last update December 2, 2024
Defined In https://doi.org/10.48550/arXiv.2304.03659
Author Madeline Chantry Schiappa
More Authors
Michael Cogswell
Ajay Divakaran
Yogesh S. Rawat
Homepage https://tinyurl.com/vlm-robustness