Datasets:

owl-owl
/

POVBench

Tasks:

Visual Question Answering

Image-Text-to-Text

Modalities:

Formats:

Languages:

Size:

Tags:

contextual-observer-grounding

spatial-reasoning

perspective-taking

vision-language-models

Libraries:

License:

Dataset card Data Studio Files Files and versions

535 MB

Ctrl+K

Ctrl+K

1 contributor

History: 13 commits

owl-owl's picture

Unify venue wording

fbaf9d7 verified 26 days ago