Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

CNN-Based Classification of Persian Miniature Paintings from Five Renowned Schools

Mojtaba Shahi, Roozbeh Rajabi, Farnaz Masoumzadeh

TL;DR

This research contributes significantly to digital art analysis, providing detailed insights into the dataset, CNN architecture, training, and validation processes, thereby aiding the preservation and understanding of Persian cultural heritage.

Abstract

This article addresses the gap in computational painting analysis focused on Persian miniature painting, a rich cultural and artistic heritage. It introduces a novel approach using Convolutional Neural Networks (CNN) to classify Persian miniatures from five schools: Herat, Tabriz-e Avval, Shiraz-e Avval, Tabriz-e Dovvom, and Qajar. The method achieves an average accuracy of over 91%. A meticulously curated dataset captures the distinct features of each school, with a patch-based CNN approach classifying image segments independently before merging results for enhanced accuracy. This research contributes significantly to digital art analysis, providing detailed insights into the dataset, CNN architecture, training, and validation processes. It highlights the potential for future advancements in automated art analysis, bridging machine learning, art history, and digital humanities, thereby aiding the preservation and understanding of Persian cultural heritage.

CNN-Based Classification of Persian Miniature Paintings from Five Renowned Schools

TL;DR

This research contributes significantly to digital art analysis, providing detailed insights into the dataset, CNN architecture, training, and validation processes, thereby aiding the preservation and understanding of Persian cultural heritage.

Abstract

This article addresses the gap in computational painting analysis focused on Persian miniature painting, a rich cultural and artistic heritage. It introduces a novel approach using Convolutional Neural Networks (CNN) to classify Persian miniatures from five schools: Herat, Tabriz-e Avval, Shiraz-e Avval, Tabriz-e Dovvom, and Qajar. The method achieves an average accuracy of over 91%. A meticulously curated dataset captures the distinct features of each school, with a patch-based CNN approach classifying image segments independently before merging results for enhanced accuracy. This research contributes significantly to digital art analysis, providing detailed insights into the dataset, CNN architecture, training, and validation processes. It highlights the potential for future advancements in automated art analysis, bridging machine learning, art history, and digital humanities, thereby aiding the preservation and understanding of Persian cultural heritage.

Paper Structure

This paper contains 12 sections, 6 figures, 5 tables.

Table of Contents

  1. Introduction
  2. Related Works
  3. Methodology
  4. Patch Extraction
  5. Transfer Learning
  6. Classification
  7. EXPERIMENTAL SETUP
  8. Dataset
  9. Classification Model
  10. Evaluation Metrics
  11. RESULTS AND DISCUSSION
  12. CONCLUSION

Figures (6)

  • Figure 1: An exquisite collage of Persian miniatures, showcasing a rich tapestry of styles from various schools of art, each piece a testament to the diverse cultural narratives and artistic expressions of the Persian heritage.
  • Figure 2: An illustration of the Qizilbash hat within the Persian miniatures of the Tabriz School.
  • Figure 3: An illustration of artwork segmented into five normalized patches, showing four non-overlapping corners and a central patch overlapping each by a quarter.
  • Figure 4: Illustrative overview of five schools of art used in this project: (a) Herat School; (b) Shiraz-e Avval School; (c) Tabriz-e Avval School; (d) Tabriz-e Dovvom School; (e) Qajar School.
  • Figure 5: Normalized Confusion Matrix of Patch-Level Classification using DenseNet201: Mean Across 5-Fold Cross-Validation.
  • ...and 1 more figures