jonathan-roberts1/CLRS
收藏Hugging Face2023-03-31 更新2024-03-04 收录
下载链接:
https://hf-mirror.com/datasets/jonathan-roberts1/CLRS
下载链接
链接失效反馈官方服务:
资源简介:
数据集CLRS是一个用于遥感图像场景分类的持续学习基准测试数据集。它包含图像和标签两个特征,标签涵盖了25种不同的场景类别,如机场、裸地、海滩等。数据集分为训练集,包含15000个样本,总大小为2969926932字节。该数据集适用于图像分类和零样本图像分类任务,且仅用于学术目的。
The CLRS dataset is a continual learning benchmark dataset for remote sensing image scene classification. It includes two features: images and labels. The labels cover 25 distinct scene categories, such as airport, bare land, beach, and so on. The dataset is split into a training set that contains 15,000 samples, with a total size of 2,969,926,932 bytes. This dataset is applicable to image classification and zero-shot image classification tasks, and is only for academic purposes.
提供机构:
jonathan-roberts1
原始信息汇总
数据集概述
数据集特征
- image: 图像数据类型。
- label: 类别标签,包含以下类别:
- 0: airport
- 1: bare land
- 2: beach
- 3: bridge
- 4: commercial
- 5: desert
- 6: farmland
- 7: forest
- 8: golf course
- 9: highway
- 10: industrial
- 11: meadow
- 12: mountain
- 13: overpass
- 14: park
- 15: parking
- 16: playground
- 17: port
- 18: railway
- 19: railway station
- 20: residential
- 21: river
- 22: runway
- 23: stadium
- 24: storage tank
数据集划分
- train: 训练集,包含15000个样本,总大小为2969926932字节。
数据集大小
- 下载大小: 2327956775字节
- 数据集大小: 2969926932字节
许可信息
- license: 其他,仅供学术使用。
任务类别
- image-classification
- zero-shot-image-classification
搜集汇总
数据集介绍

背景与挑战
背景概述
CLRS数据集是一个用于遥感图像场景分类的持续学习基准数据集,包含15,000张256x256像素的图像,涵盖25个类别标签。该数据集支持图像分类和零样本图像分类任务,适用于学术研究用途。
以上内容由遇见数据集搜集并总结生成



