ImagenHub/Text_Guided_Image_Editing

---
language:
- en
license: cc-by-4.0
size_categories:
- n<1K
task_categories:
- image-to-image
dataset_info:
features:
- name: img_id
dtype: string
- name: turn_index
dtype: int32
- name: source_img
dtype: image
- name: mask_img
dtype: image
- name: instruction
dtype: string
- name: source_global_caption
dtype: string
- name: target_global_caption
dtype: string
- name: target_local_caption
dtype: string
- name: target_img
dtype: image
splits:
- name: dev
num_bytes: 1521276668.0
num_examples: 528
- name: filtered
num_bytes: 504007147.0
num_examples: 179
- name: extra
num_bytes: 709468665.0
num_examples: 249
download_size: 2734685875
dataset_size: 2734752480.0
configs:
- config_name: default
data_files:
- split: dev
path: data/dev-*
- split: filtered
path: data/filtered-*
- split: extra
path: data/extra-*
---
# Dataset Card

Dataset in [ImagenHub](arxiv.org/abs/2310.01596).

# Citation

Please kindly cite our paper if you use our code, data, models or results:

```
@article{ku2023imagenhub,
title={ImagenHub: Standardizing the evaluation of conditional image generation models},
author={Max Ku and Tianle Li and Kai Zhang and Yujie Lu and Xingyu Fu and Wenwen Zhuang and Wenhu Chen},
journal={arXiv preprint arXiv:2310.01596},
year={2023}
}
```

原始信息汇总

数据集概述

基本信息

• 语言: 英语
• 许可证: CC-BY-4.0
• 大小分类: 小于1K
• 任务分类: 图像到图像

数据集特征

• img_id: 字符串类型
• turn_index: 整数类型
• source_img: 图像类型
• mask_img: 图像类型
• instruction: 字符串类型
• source_global_caption: 字符串类型
• target_global_caption: 字符串类型
• target_local_caption: 字符串类型
• target_img: 图像类型

数据集分割

• dev: 528个样本，1521276668字节
• filtered: 179个样本，504007147字节
• extra: 249个样本，709468665字节

下载与数据集大小

• 下载大小: 2734685875字节
• 数据集大小: 2734752480字节

配置文件

• 默认配置:
  • dev: 数据路径为data/dev-*
  • filtered: 数据路径为data/filtered-*
  • extra: 数据路径为data/extra-*