High-throughput computational stylometry and quantitative connoisseurship engine for fine art paintings: patch-level micro-texture analysis, self-supervised vision embeddings (DINOv2), multi-hand intra-canvas attribution heatmaps, and outlier clustering via IIIF museum archives.
computer-vision digital-humanities iiif anomaly-detection hdbscan art-history unsupervised-clustering texture-analysis fine-art brushstroke-analysis computational-stylometry dinov2 patch-embeddings connoisseurship art-attribution
-
Updated
Oct 26, 2026 - HTML