23 datasets found

A dataset for fill-in-the-blank tasks, including video question answering.

NExT-GQA is an extension of NExT-QA with 10.5K temporal grounding (or location) labels tied to the original QA pairs.

Video question answering (VideoQA) is an essential task in vision-language understanding, which has attracted numerous research attention recently. Nevertheless, existing works...

The STAR dataset provides 60K situated reasoning questions based on 22K trimmed situation video clips.

The AGQA benchmark is a visual dataset comprising 192M hand-crafted questions about 9.6K videos from the Charades dataset.

The SHG-VQA model predicts a situation hyper-graph structure composed of existing actions and relations in the input video.

The HealthVidQA-Prompt dataset is a large-scale medical instructional video question-answering dataset. It contains 52,771 video-question-answer triplets from 13,990 medical...

The HealthVidQA-CRF dataset is a large-scale medical instructional video question-answering dataset. It contains 23,434 video-question-answer triplets from 11,708 medical videos.

A video story question answering dataset containing 24,282 questions about 207 episodes of The Big Bang Theory.

A dataset for video question answering, including video evidence selection.

Progressive Graph Attention Network for Video Question Answering.

TVQA is a video question answering dataset collected from 6 long-running TV shows from 3 genres. There are 21,793 video clips in total for QA collection, accompanied with...

This paper proposes a video question answering model that effectively integrates multi-modal input sources and finds the temporally relevant information to answer questions.

Video question answering (VideoQA) requires systems to understand the visual information and infer an answer for a natural language question from it.

Video-Language Models (VLMs), powered by the advancements in Large Language Models (LLMs), are charting new frontiers in video understanding. A pivotal challenge is the...

Video question answering task, which requires machines to answer questions about videos in a natural language form.

The MSVD-QA dataset is a benchmark for video question answering, containing 1,970 videos with multiple-choice questions.

The TGIF-QA dataset consists of 165165 QA pairs chosen from 71741 animated GIFs. To evaluate the spatiotemporal reasoning ability at the video level, TGIF-QA dataset designs...

Zero-shot video question answering via frozen bidirectional language models.

The dataset used in the paper is a video question answering dataset, which is a large-scale video-language pre-training task.