pemujo/GLDv2_Top_51_Categories

---
dataset_info:
features:
- name: id
dtype: string
- name: landmark_id
dtype: int64
- name: category
dtype: string
- name: image
dtype: image
- name: label
dtype: int64
splits:
- name: train
num_bytes: 2428986323.125
num_examples: 36463
- name: test
num_bytes: 606874794.5
num_examples: 9116
download_size: 3034360629
dataset_size: 3035861117.625
language:
- en
pretty_name: GLDv2 Top 51 Categories
size_categories:
- n<1K
---
# Dataset Card for Dataset Name

### Dataset Summary

This dataset is a subset of Kaggle's Google Landmark Recognition 2021 competition with only the categories with more than 500 images.
https://www.kaggle.com/competitions/landmark-recognition-2021/data

The dataset consists of a total of 45579 224x224 color images in 51 categories.

### Languages

English

## Dataset Structure

### Data Fields

- `landmark_id`: Int - Numeric identifier of the category
- `category` : String - Name of the category
- `id` : String - Image identifier
- `image` : Image - PIL image object
- `label` : Int - Numeric label from 0 to 50

### Data Splits

The dataset was randomly split with 80% of the images for the train set and 20% for the test set.

| | train | test |
|----------------------|------:|-----:|
| Dataset | 36463 | 9116 |

### Source Data

The full dataset is from Kaggle Landmark Recognition 2021

"Towards A Fairer Landmark Recognition Dataset", Z. Kim, A. Araujo, B. Cao, C. Askew, J. Sim, M. Green, N. Yilla and T. Weyand, arxiv:2108.08874

https://www.kaggle.com/competitions/landmark-recognition-2021/data

### Citation Information

"Google Landmarks Dataset v2 - A Large-Scale Benchmark for Instance-Level Recognition and Retrieval", T. Weyand, A. Araujo, B. Cao and J. Sim, Proc. CVPR'20

"Towards A Fairer Landmark Recognition Dataset", Z. Kim, A. Araujo, B. Cao, C. Askew, J. Sim, M. Green, N. Yilla and T. Weyand, arxiv:2108.08874

原始信息汇总

数据集概述

• 名称: GLDv2 Top 51 Categories
• 数据集大小: 总大小为3035861117.625字节
• 下载大小: 3034360629字节
• 语言: 英语
• 类别数量: 51
• 图像总数: 45579张
• 图像尺寸: 224x224像素，彩色图像
• 数据分割:
  • 训练集: 36463张图像，占80%
  • 测试集: 9116张图像，占20%

数据字段

• id: 字符串类型，图像标识符
• landmark_id: 整数类型，类别数字标识
• category: 字符串类型，类别名称
• image: 图像类型，PIL图像对象
• label: 整数类型，数值标签，范围从0到50

数据集来源

• 原始数据集: Kaggle Landmark Recognition 2021
• 相关研究:
  • "Google Landmarks Dataset v2 - A Large-Scale Benchmark for Instance-Level Recognition and Retrieval", T. Weyand, A. Araujo, B. Cao and J. Sim, Proc. CVPR20
  • "Towards A Fairer Landmark Recognition Dataset", Z. Kim, A. Araujo, B. Cao, C. Askew, J. Sim, M. Green, N. Yilla and T. Weyand, arxiv:2108.08874