Skip to content

TVQA

Multi-Modal LearningVideo Question AnsweringEnglish

TVQA is a multi-modal learning dataset in English from Lei et al. with 460+ Hours records in HDF5, JSON format.

About TVQA

Dataset is used for video question answering and consists of 152,545 QA pairs from 21,793 clips, spanning over 460 hours of video.

Details

Task
Multi-Modal Learning, Video Question Answering
Language
English
Format
HDF5, JSON
Rows / instances
460+ Hours
Creator
Lei et al.
Year
2018
Download Paper

Related Multi-Modal Learning, Video Question Answering datasets

FAQ