Clustering Artworks by Ai-Quantified Visual Qualities & Content Recommendation App Project by Gabriel del Valle, published October 8 2024 For any questions about this project please feel free to reach out on Linkedin: www.linkedin.com/in/gabrielxdelvalle
https://nycdatascience.com/blog/student-works/clustering-artworks-by-ai-quantified-visual-qualities-content-recommendation-app/
This project contains 3 sequential notebooks with detailed notes and comments 01_artvee_art_webscrape : Python Jupyter Notebook : Webscraping artworks, cleaning data, creating image links to host images online using github 02_generate_categories : Python Jupyter Notebook : Generate image scores for artworks using CLIP and store as dataframes 03_evaluate_categories : R Markdown Notebook : Evaluate effectiveness of image score categories, fit cluster models, graph metrics for:
Within Sum of Squares
Between Sum of Squares / Sum of Squares
Silhouette Scores
PCA Graphs
app.R is a shiny app, launched to the shiny.io server using RSconnect colors_gallery.csv is the dataset used by app.R https://gabrielxdelvalle.shinyapps.io/algo_gallery/