Audio Visual Scene-Aware Dialog (AVSD)
Multi-Modal LearningVideo Question AnsweringDialogueEnglish
Audio Visual Scene-Aware Dialog (AVSD) is a multi-modal learning-focused dataset in English that provides 11,816 labeled examples distributed in JSON format.
About Audio Visual Scene-Aware Dialog (AVSD)
Dataset consists of text-based human conversations about short videos from the Charades dataset.
Details
- Task
- Multi-Modal Learning, Video Question Answering, Dialogue
- Language
- English
- Format
- JSON
- Rows / instances
- 11,816
- Creator
- Alamri et al.
- Year
- 2019