AVSD|对话理解数据集|视听数据处理数据集

displayName: AVSD (Audio-Visual Scene-Aware Dialog)
license:
- MIT
mediaTypes:
- Audio
paperUrl: https://arxiv.org/pdf/1806.00525v1.pdf
publishDate: "2019"
publishUrl: http://workshop.colips.org/dstc7/call.html
publisher:
- Mitsubishi Electric Research Laboratories
- Georgia Institute of Technology
tags:
- Visual Question-Answer
taskTypes:
- Scene-Aware Dialogue

---
# 数据集介绍
## 简介
视听场景感知对话 (AVSD) 数据集，或DSTC7轨道3，是用于对话理解的视听数据集。数据集和轨道的目标是设计系统，以在给定视频的对话历史和视听内容的情况下，在有关视频的对话框中生成响应。
## 引文
```
@article{alamri2018audio,
title={Audio visual scene-aware dialog (avsd) challenge at dstc7},
author={Alamri, Huda and Cartillier, Vincent and Lopes, Raphael Gontijo and Das, Abhishek and Wang, Jue and Essa, Irfan and Batra, Dhruv and Parikh, Devi and Cherian, Anoop and Marks, Tim K and others},
journal={arXiv preprint arXiv:1806.00525},
year={2018}
}
```

## Download dataset
:modelscope-code[]{type="git"}