WildQA is a video understanding dataset of videos recorded in outside settings. The dataset can be used to evaluate models for video question answering.
Variants: WildQA
This dataset is used in 1 benchmark:
| Task | Model | Paper | Date |
|---|---|---|---|
| Video Question Answering | Multi (text + video, IO) | WildQA: In-the-Wild Video Question Answering | 2022-09-14 |
| Video Question Answering | Multi (text + video, SE) | WildQA: In-the-Wild Video Question Answering | 2022-09-14 |
| Video Question Answering | T5 (text) | WildQA: In-the-Wild Video Question Answering | 2022-09-14 |
| Video Question Answering | T5 (text + video) | WildQA: In-the-Wild Video Question Answering | 2022-09-14 |
| Video Question Answering | T5 (text, zero-shot) | WildQA: In-the-Wild Video Question Answering | 2022-09-14 |
Recent papers with results on this dataset: