WikiArtVectors: Style and Color Representations of Artworks for Cultural Analysis via Information Theoretic Measures
Why this work is in the frame
A frame that forgets how it found something cannot be audited. These are the routes that admitted this work.
Bibliographic record
Abstract
With the increase in massive digitized datasets of cultural artefacts, social and cultural scientists have an unprecedented opportunity for the discovery and expansion of cultural theory. The WikiArt dataset is one such example, with over 250,000 high quality images of historically significant artworks by over 3000 artists, ranging from the 15th century to the present day; it is a rich source for the potential mining of patterns and differences among artists, genres, and styles. However, such datasets are often difficult to analyse and use for answering complex questions of cultural evolution and divergence because of their raw formats as image files, which are represented as multi-dimensional tensors/matrices. Recent developments in machine learning, multi-modal data analysis and image processing, however, open the door for us to create representations of images that extract important, domain-specific features from images. Art historians have long emphasised the importance of art style, and the colors used in art, as ways to characterise and retrieve art across genre, style, and artist. In this paper, we release a massive vector-based dataset of paintings (WikiArtVectors), with style representations and color distributions, which provides cultural and social scientists with a framework and database to explore relationships across these two vital dimensions. We use state-of-the-art deep learning and human perceptual color distributions to extract the representations for each painting, and aggregate them across artist, style, and genre. These vector representations and distributions can then be used in tandem with information-theoretic and distance metrics to identify large-scale patterns across art style, genre, and artist. We demonstrate the consistency of these vectors, and provide early explorations, while detailing future work and directions. All of our data and code is publicly available on GitHub.
Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.
Full frame distilled prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. Learned from the 10,348 direct Codex labels and 10,348 direct Gemma labels. Candidate is the union of thresholded teacher heads; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels or direct frontier model labels.
Codex and Gemma teacher scores by category
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.000 | 0.000 |
| Meta-epidemiology (narrow) | 0.000 | 0.000 |
| Meta-epidemiology (broad) | 0.000 | 0.000 |
| Bibliometrics | 0.000 | 0.001 |
| Science and technology studies | 0.000 | 0.000 |
| Scholarly communication | 0.000 | 0.000 |
| Open science | 0.000 | 0.000 |
| Research integrity | 0.000 | 0.000 |
| Insufficient payload (model declined to judge) | 0.000 | 0.000 |
Machine scores (provisional)
The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.
Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.
score_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from it