Recap-COCO-30K

## Llava recaptioned COCO2014 ValSet.

Used for text-to-image generation evaluaion. More detial can be found in What If We Recaption Billions of Web Images with LLaMA-3?

## Dataset Structure
<!-- This section provides a description of the dataset fields, and additional information about the dataset structure such as criteria used to create the splits, relationships between data points, etc. -->
"image_id" (str): COCO image id.

"coco_url" (image): the COCO image url.

"caption" (str): the original COCO caption.

"recaption" (str): the llava recaptioned COCO caption.

## Citation

<!-- If there is a paper or blog post introducing the dataset, the APA and Bibtex information for that should go in this section. -->

**BibTeX:**

```
@article{li2024recapdatacomp,
title={What If We Recaption Billions of Web Images with LLaMA-3?},
author={Li, Xianhang and Tu, Haoqin and Hui, Mude and Wang, Zeyu and Zhao, Bingchen and Xiao, Junfei and Ren, Sucheng and Mei, Jieru and Liu, Qing and Zheng, Huangjie and Zhou, Yuyin and Xie, Cihang},
journal={arXiv preprint arXiv:2406.12345},
year={2024}
```

Llava重标注COCO2014验证集（Llava recaptioned COCO2014 ValSet）。
本数据集用于文本到图像生成（text-to-image generation）任务的评估。更多细节可参阅论文《What If We Recaption Billions of Web Images with LLaMA-3?》。

## 数据集结构
本节提供数据集字段说明，以及数据集划分依据、数据点间关联关系等数据集结构相关额外信息：
- `image_id`（字符串类型）：COCO图像编号。
- `coco_url`（图像链接）：COCO数据集官方图像访问URL。
- `caption`（字符串类型）：原始COCO图像标注文本。
- `recaption`（字符串类型）：经Llava重生成的COCO图像标注文本。

## 引用
**BibTeX：**

@article{li2024recapdatacomp,
title={What If We Recaption Billions of Web Images with LLaMA-3?},
author={Li, Xianhang and Tu, Haoqin and Hui, Mude and Wang, Zeyu and Zhao, Bingchen and Xiao, Junfei and Ren, Sucheng and Mei, Jieru and Liu, Qing and Zheng, Huangjie and Zhou, Yuyin and Xie, Cihang},
journal={arXiv preprint arXiv:2406.12345},
year={2024}
}

搜集汇总

背景与挑战

背景概述

Recap-COCO-30K是一个用于文本到图像生成评估的数据集，基于COCO2014验证集，通过LLaMA-3对原始标题进行了重新标注，包含图像ID、URL、原始标题和重新标注的标题。