---
title: CUB-200-2011
canonical_url: "https://www.modelscope.cn/datasets/OpenDataLab/CUB-200-2011"
md_url: "https://www.modelscope.cn/datasets/OpenDataLab/CUB-200-2011.md"
repository: OpenDataLab/CUB-200-2011
last_updated: 2024-07-08
license: "[CUB-200-2011 Custom]"
storage_size: "1.1 GB"
domain:
  - publisher
  - labelTypes
  - displayName
  - publishUrl
  - paperUrl
  - mediaTypes
  - taskTypes
tasks:
  - "Allan Lab"
  - Classification
  - "CUB-200-2011 (Caltech-UCSD Birds-200-2011)"
  - "http://www.vision.caltech.edu/datasets/cub_200_2011/"
  - "https://arxiv.org/abs/1709.00340"
  - Image
  - "Object Detection"
downloads: 1507
stars: 3
---

# CUB-200-2011

> CUB-200-2011 - OpenDataLab 在 ModelScope 开源的数据集。displayName: CUB-200-2011 (Caltech-UCSD Birds-200-2011) labelTypes: Multi Attributes Classification Classification license: CUB-200-2011 Custom mediaTypes: Image paperUrl: https://arxiv.org/abs/1709.00340…

OpenDataLab/CUB-200-2011 是 ModelScope 魔搭社区上的Allan Lab、Classification、CUB-200-2011 (Caltech-UCSD Birds-200-2011)数据集，涉及 publisher、labelTypes、displayName 领域，存储大小 1.1 GB，采用 [CUB-200-2011 Custom] 许可。

- **Repository**: OpenDataLab/CUB-200-2011
- **License**: [CUB-200-2011 Custom]
- **Tasks**: Allan Lab, Classification, CUB-200-2011 (Caltech-UCSD Birds-200-2011), http://www.vision.caltech.edu/datasets/cub_200_2011/, https://arxiv.org/abs/1709.00340, Image, Object Detection
- **Domain**: publisher, labelTypes, displayName, publishUrl, paperUrl, mediaTypes, taskTypes
- **Storage size**: 1.1 GB
- **Downloads**: 1507
- **Stars**: 3
- **Last updated**: 2024-07-08

Source: https://www.modelscope.cn/datasets/OpenDataLab/CUB-200-2011

---

displayName: CUB-200-2011 (Caltech-UCSD Birds-200-2011)
labelTypes:
- Multi Attributes Classification
- Classification
license:
- CUB-200-2011 Custom
mediaTypes:
- Image
paperUrl: https://arxiv.org/abs/1709.00340
publishDate: ""
publishUrl: http://www.vision.caltech.edu/datasets/cub_200_2011/
publisher:
- Allan Lab
tags:
- Bird
taskTypes:
- Image Classification
- Fine Grained Image Classification
- Attribute-based Image Recognition
- Part Localization
- Object Detection

---
# 数据集介绍
  ## 简介
  Caltech-UCSD Birds-200-2011 (CUB-200-2011) 数据集是用于细粒度视觉分类任务的最广泛使用的数据集。它包含属于鸟类的 200 个子类别的 11,788 张图像，其中 5,994 张用于训练，5,794 张用于测试。每个图像都有详细的注释：1 个子类别标签、15 个部分位置、312 个二进制属性和 1 个边界框。文本信息来自 Reed 等人。他们通过收集细粒度的自然语言描述来扩展 CUB-200-2011 数据集。为每个图像收集十个单句描述。自然语言描述是通过 Amazon Mechanical Turk (AMT) 平台收集的，要求至少 10 个单词，没有任何子类别和动作信息。
  ## 引文
  ```
@article{wah2011caltech,
  title={The caltech-ucsd birds-200-2011 dataset},
  author={Wah, Catherine and Branson, Steve and Welinder, Peter and Perona, Pietro and Belongie, Serge},
  year={2011},
  publisher={California Institute of Technology}
}
```
  
## Download dataset
:modelscope-code[]{type="git"}