I'm trying to use IG-3.5B-17k dataset from paperswithcode.com, but I can't figure out how to do that exactly. How to do it? I guess I need to use this site API? Am I even able to use this dataset?


1 Answer 1


This is specified in the dataset description in paperswithcode.com (emphasis mine):

IG-3.5B-17k is an internal Facebook AI Research dataset for training image classification models. It consists of hashtags for up to 3.5 billion public Instagram images.

So the dataset is not public.

In the paper, however, the authors argue that the images and their hashtags are "visible" in Instagram:

Our datasets have two nice properties: public visibility and simplicity. By using publicly accessible images, the data used in our experiments is visible to everyone. To see what it looks like, the images are browsable by hashtag at https://www.instagram.com/explore/tags/ followed by a specific hashtag; for example https://www.instagram.com/explore/tags/brownbear shows images tagged with #brownbear. Our data is also taken from the “wild”, essentially as-is, with minimal effort to sanitize it. This makes the dataset construction process particularly simple and transparent.

I understand that they are saying "you can go and query Instagram yourself" to see the images in the dataset, but I don't think this is actually practical or even allowed in their terms of service.


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.