DVQA
Question AnsweringVisualCommonsenseEnglish
DVQA is a question answering dataset in English from Kafle et al. with 3,487,194 records in JSON, PNG format.
Details
- Task
- Question Answering, Visual, Commonsense
- Language
- English
- Format
- JSON, PNG
- Rows / instances
- 3,487,194
- Creator
- Kafle et al.
- Year
- 2018