Data Science for Measuring Visual Influence in Art

Bibliographic Data

ID22197663
AuthorsNittin Sharma (0000-0002-8440-8374, Chitkara University), Mona Sharma (0000-0002-4854-1725, Noida International University), Pratik Mungekar (Society for Applied Microwave Electronics Engineering & Research), Aswathy N Rajan (Presidency University), Omprakash Das (Siksha O Anusandhan University), Shivam Khurana (Chitkara University)
Year2025
Volume6
Issue2s
Publication date2025-12-16
Open AccessYes
TypeARTICLE
VenueShodhKosh: Journal of Visual and Performing Arts (JOURNAL)
Journal identifiersISSN: 2582-7472 • E-ISSN: 2582-7472
PublisherGranthaalayah Publications and Printers (PUBLISHER • IN)
DOI10.29121/shodhkosh.v6.i2s.2025.6744
OpenAlexW4417458678
LanguageEN
References cited17

The paper discusses the applications of data-driven approaches to quantify visual influence in art which combines computational means with art-historical theory to identify stylistic and conceptual connections between artists and works of art. The paper investigates the way data science can give objective schemes of the artistic evolution by quantifying the visual similarities and the pathways of influence. Based on online art repositories massive datasets, preprocessing and transformation of images is done to a multi-dimensional representation of features that reflect style, color, texture, and composition. A number of machine learning models, including Convolutional Neural Networks (CNNs), Variational Autoencoders (VAEs), Graph Neural Networks (GNNs) and the K-Means clustering, are used to identify latent stylistic patterns, as well as build influence networks through time, per movement. These computational mappings are compared with the network theory to compute visual lineages and make inferences on the likely channels of art transmission. Although the method improves the level of empirical rigor in influence studies, the study has noted several limitations: the subjectivity of defining influence, the tendency of focusing more on the visual aspects than conceptual aspects, and the inability to establish the level of ground truth for validation

Artificial neural network · Convolutional neural network · Deep learning · Graph · Graphical model · Preprocessor · Visual reasoning · Visualization · Aesthetic Perception and Analysis · Art History and Market Analysis · Generative Adversarial Networks and Image Synthesis

Tools

Open DOIOpen Access