Skip to content

Latest commit

 

History

History
12 lines (9 loc) · 448 Bytes

File metadata and controls

12 lines (9 loc) · 448 Bytes

DramaQA+K dataset for Learning by Asking

Extends DramaQA, the existing benchmark VideoQA dataset. DramaQA+K dataset is structured hierarchically based on video length: Episode, Scene, Shot Each level includes captions, knowledge graphs, and addtional question-answer pairs.

Download dataset

git clone https://github.com/cws7777/lba_datasets.git

image