VQA-Introspect
Question AnsweringVisualEnglish
VQA-Introspect is a question answering-focused dataset in English that provides 238 labeled examples distributed in JSON format.
About VQA-Introspect
Dataset consists of 238K new perception questions from the VQA dataset which serve as sub questions corresponding to the set of perceptual tasks needed to answer complex reasoning questions.
Details
- Task
- Question Answering, Visual
- Language
- English
- Format
- JSON
- Rows / instances
- 238
- Creator
- Selvaraju et al.
- Year
- 2020