[Corpora-List] Corpora for an automatic image annotation model

[Dan Smith]

Dear members,

I have been having some trouble trying to find a dataset / test collection / corpus with the following three pieces of complementary data:

- An image (i)
- The image (i) annotations (a)
- Auxiliary/complementary text (t) associated with the image (i)

An example of one piece of test data would be similar to a news article (e.g. from the BBC website), but one that includes image annotations.

I have only been able to find datasets of (images+annotations), (images+blob regions+annotations) and so on...Ideally I'm looking for (images+annotations+text).

It is important that there is a certain "body" text to be processed.

Any help is greatly appreciated,
Dan Smith
_______________________________________________
Corpora mailing list
Corpora at uib.no
http://mailman.uib.no/listinfo/corpora