mirror of
https://github.com/openai/CLIP.git
synced 2026-11-05 01:37:26 +00:00
[GH-ISSUE #154] Availability of Rendered SST2 and Country211 datasets #134
Labels
No labels
pull-request
No milestone
No project
No assignees
1 participant
Notifications
Due date
No due date set.
Dependencies
No dependencies set.
Reference
openai/CLIP#134
Loading…
Add table
Add a link
Reference in a new issue
No description provided.
Delete branch "%!s()"
Deleting a branch is permanent. Although the deleted branch may continue to exist for a short time before it actually gets removed, it CANNOT be undone in most cases. Continue?
Originally created by @pj-ms on GitHub (Sep 16, 2021).
Original GitHub issue: https://github.com/openai/CLIP/issues/154
Hi,
Thanks for the great work! In the paper, it mentions "we created two datasets that we call Country211 and Rendered SST2. The Country211 dataset is designed to assess the geolocation capability of visual representations. We filtered the YFCC100m dataset (Thomee et al., 2016) to find 211 countries (defined as having an ISO-3166 country code) that have at least 300 photos with GPS coordinates, and we built a balanced dataset with 211 categories, by sampling 200 photos for training and 100 photos for testing, for each country"
I wonder if it is possible to share those two datasets for reproducibility purpose? Thanks!
@fuersta commented on GitHub (Sep 20, 2021):
Hi,
i would also be interested in these datasets. Especially Country211 as i find this type of problem really interesting.
Thanks!
@jongwook commented on GitHub (Sep 24, 2021):
Added in
efe8cbbdf3!@pj-ms commented on GitHub (Sep 24, 2021):
@jongwook Thanks for adding them!
@pj-ms commented on GitHub (Sep 24, 2021):
Hi, @jongwook, btw, in CLIP paper, the number of training images is listed as 43200, which is probably a typo, as 200 * 211 = 42200.