metadata
license: mit
dataset_info:
features:
- name: image
dtype: image
- name: label
dtype:
class_label:
names:
'0': Altocumulus
'1': Altostratus
'2': Cirrocumulus
'3': Cirrostratus
'4': Cirrus
'5': Cumulonimbus
'6': Cumulus
'7': Nimbostratus
'8': Stratocumulus
'9': Stratus
splits:
- name: train
num_bytes: 395771309
num_examples: 251
download_size: 395107698
dataset_size: 395771309
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
CCAiM CloudsDataset
Description
This dataset contains photographs of clouds collected for the CCAiM project, a model for cloud classification. It includes various types of clouds captured from the ground and can be used for training and testing computer vision models.
Dataset Structure
- Cloud images in JPEG/PNG format
- Optional metadata: cloud type, date, location
Usage
all data
from datasets import load_dataset
dataset = load_dataset("serbekun/CCAiM-CloudsDataset")
print(dataset)
Only one file
from datasets import load_dataset
ds = load_dataset("serbekun/CCAiM-CloudsDataset")
example = ds["train"][0]
image = example["image"]
label = example["label"]
image.show()
print("Label:", ds["train"].features["label"].int2str(label))
You will get a DatasetDict with splits like train, validation (if available) and images of clouds with their corresponding labels.
Project Purpose
The dataset is intended for training and testing the CCAiM model for cloud classification. It can be used for educational and research purposes.
License
The dataset is released under the MIT
Contact
For questions or suggestions, reach out via GitHub: serbekun