Dant33/WikiArt-81K-BLIP_2-768x768

WikiArt Resized Dataset

Description

This dataset contains 81,444 artistic images from WikiArt, organized into different artistic genres. The images have been resized to a uniform resolution of 768x768 pixels using LANCZOS resampling and padded to maintain aspect ratio, ensuring consistency for machine learning tasks and computational art analysis. The base for the dataset was Dant33/WikiArt-81K-BLIP_2-captions.

Enhancements

1. Image Resizing with Padding

• All images have been resized to fit within 768x768 pixels while maintaining their aspect ratio.
• The resizing was performed using target_size=(768, 768), resample=Image.LANCZOS, followed by padding with a black background to fill the remaining space.
• The filenames were updated to reflect this change by appending _resize768 to the original filename.

2. Genre Normalization

• Each image is classified under a single main artistic genre.
• The dataset contains 27 main genres:

#	Genre	Size
1	Impressionism	13060
2	Realism	10733
3	Romanticism	7019
4	Expressionism	6736
5	Post Impressionism	6450
6	Symbolism	4528
7	Art Nouveau Modern	4334
8	Baroque	4240
9	Abstract Expressionism	2782
10	Northern Renaissance	2552
11	Naive Art Primitivism	2405
12	Cubism	2235
13	Rococo	2089
14	Color Field Painting	1615
15	Pop Art	1483
16	Early Renaissance	1391
17	High Renaissance	1343
18	Minimalism	1337
19	Mannerism Late Renaissance	1279
20	Ukiyo e	1167
21	Fauvism	934
22	Pointillism	513
23	Contemporary Realism	481
24	New Realism	314
25	Synthetic Cubism	216
26	Analytical Cubism	110
27	Action painting	98

3. Automatic Description Generation

• Each image includes a textual description generated using the BLIP2 model.

Dataset Format

The dataset is structured in a CSV file with the following columns:

• file_name: Image filename, including the _resize768 suffix.
• genre: Assigned artistic genre.
• artist: Artist's name.
• painting_name: Painting title, cleaned and formatted.
• phash: Perceptual hash of the image.
• description: Automatically generated image description.
• subset: Subset to which the image belongs (train or test).

Example Entry

file_name,genre,artist,painting_name,phash,description,subset
Impressionism/william-merritt-chase_still-life-with-cockatoo_resize768.jpg,Impressionism,william merritt chase,still life with cockatoo,b0e24b85961e6de9,a painting of a white bird sitting on a vase,train

Usage

This dataset is ideal for:

• Artistic style classification.
• Training deep learning models for image-based tasks.
• Automatic art description generation.
• Image retrieval and similarity analysis.