Divya0001/indian-art-styles

🎨 Indian Art Styles Dataset

A comprehensive image classification dataset covering 34 distinct Indian painting and art styles with 27,139 images in total. This dataset is designed for training Vision Transformer (ViT) and CNN-based classifiers to recognize traditional Indian art styles.

Dataset Overview

Style	Region	Medium	Image Count
aipan	Uttarakhand	floor/wall painting	6
bengal_school	West Bengal	painting	1287
bhil	Madhya Pradesh / Rajasthan / Gujarat	painting	52
bikaner	Rajasthan	miniature painting	172
bundi	Rajasthan	miniature painting	185
cheriyal	Telangana	scroll painting	1292
gond	Madhya Pradesh	painting	1260
jaipur	Rajasthan	miniature painting	264
kalamkari	Andhra Pradesh	textile/painting	1291
kalighat	West Bengal	painting	2165
kangra	Himachal Pradesh	miniature painting	566
kerala_mural	Kerala	mural/wall painting	842
kishangarh	Rajasthan	miniature painting	176
kota	Rajasthan	miniature painting	178
madhubani	Bihar	painting	1257
mandana	Rajasthan	floor/wall painting	175
manjusha	Bihar	painting	7
mewar	Rajasthan	miniature painting	818
mughal_miniature	Mughal Empire (Pan-India)	miniature painting	2056
mysore	Karnataka	painting	56
nirmal	Telangana	painting	13
pahari	Himachal Pradesh / Jammu	miniature painting	1142
patna_kalam	Bihar	painting	476
pattachitra	Odisha / West Bengal	painting	2551
phad	Rajasthan	scroll painting	1297
pichwai	Rajasthan	painting	1301
pithora	Gujarat / Rajasthan / Madhya Pradesh	wall painting	292
rajput	Rajasthan	miniature painting	1145
rogan	Gujarat	fabric painting	173
saura	Odisha	wall painting	1219
sohrai	Jharkhand	wall painting	13
tanjore	Tamil Nadu	painting	964
thangka	Sikkim / Ladakh / Arunachal Pradesh	painting	1185
warli	Maharashtra	painting	1263

Total: 27,139 images across 34 styles

Dataset Structure

dataset/
├── train/          # Training images organized by style
│   ├── aipan/
│   ├── bengal_school/
│   ├── bhil/
│   └── ... (34 style directories)
└── metadata.csv    # Full manifest with style, source, dimensions, hash, URLs

Each image is:

• Minimum 400×400 pixels (permissive filter to preserve portrait/scroll formats)
• Sourced from verified museum APIs and cultural archives
• Deduplicated using file hash + URL checks

Data Sources

Images were collected from:

• Wikimedia Commons (category verification + subcategory depth traversal)
• Metropolitan Museum of Art API
• Cleveland Museum of Art API
• Victoria and Albert Museum (V&A) API
• Europeana API
• British Library IIIF
• Rubin Museum
• Web Search fallback (Google Custom Search / Bing Image Search for minority styles)

Usage

from datasets import load_dataset

# Load all styles together
ds = load_dataset("sohinipattanayak/indian-art-styles")

# Load a specific style as a subset
ds = load_dataset("sohinipattanayak/indian-art-styles", name="madhubani")

License

Images are sourced from public domain or open-license museum collections (CC BY 4.0 compatible). This dataset compilation is released under CC BY 4.0.

Citation

If you use this dataset, please cite:

@dataset{indian_art_styles_2026,
  title={Indian Art Styles Image Dataset},
  author={Pattanayak, Sohini},
  year={2026},
  publisher={Hugging Face},
  url={https://huggingface.co/datasets/sohinipattanayak/indian-art-styles}
}