Power, Control, and Data Processing Systems

Classification of Persian Miniature Painting Schools Using a Vision Transformer and Transfer Learning

Document Type : Original Research

Authors

Darya Seifi Zarei 1
HamidReza Shahdoosti 2

1 Electrical and Computer Department, Hamedan University of Technology, Hamedan, Iran.

2 Electrical and Computer Engineering Department, Hamedan University of Technology, Hamedan, Iran

https://doi.org/10.30511/pcdp.2026.2095440.1101
Abstract
The automated attribution of historical paintings to their school of origin is a long-standing challenge at the intersection of computer vision and art history. This challenge is especially acute for non-Western art traditions, such as Persian miniature painting, for which digitized and labelled image corpora remain comparatively small. This paper addresses the classification of Persian miniature paintings into five historical schools including, Herat, Qajar, Shiraz-e Avval, Tabriz-e Avval, and Tabriz-e Dovvom, using a Vision Transformer (ViT). Transfer learning is used to adapt a ViT-B/16 model, pre-trained on ImageNet-21k, to this task. Because the available dataset contains only 890 images, five complementary augmentation operations i.e. random-resized cropping, horizontal flipping, rotation, color jitter, and random erasing, are applied during training, together with Mixup as an additional sample-mixing regularizer. A progressive layer-freezing strategy keeps the first six Transformer encoder blocks fixed and fine-tunes only the last six blocks. To stabilize training on the small fine-tuning set, this strategy utilizes a classification head combined with label smoothing and a low, cosine-annealing learning rate. On a held-out test set of 178 images, the fine-tuned model reaches an overall accuracy of 92.13% and a weighted F1-score of 0.92, with three of the five schools classified at an F1-score above 0.94. These results indicate that a Vision Transformer, can reach competitive accuracy on a small, specialized cultural-heritage dataset when transfer learning and augmentation are used simultaneously.

Keywords

Vision Transformer
Persian miniature painting
transfer learning
data augmentation
image classification
digital humanities
Subjects
[1]    Huang, Q., Li, M., Agustin, D., Li, L. and Jha, M., (2023). A novel CNN model for classification of Chinese historical calligraphy styles in regular script font. Sensors, 24(1), 188-197. https://doi.org/10.3390/s24010197
[2]    Ghasemi, S., Ghasemi, S., Rahmani, A. M., & Barzamini, R. (2025). Transfer Learning-based Detection of COVID-19 Cases from Chest CT Scans. Power, Control, and Data Processing Systems, 2(3), 33-47. https://doi.org/10.30511/pcdp.2025.2063435.1030
[3]    Amirbeigi, H., Shahdoosti, H. R., (2026) A New Method for Detecting Emotions in Video Using a Hybrid CNN-RNN-RBM Network. Power, Control, and Data Processing Systems, (under publication). https://doi.org/10.30511/pcdp.2026.2087714.1069
[4]    Jamali, L., 2013. Colouring the words: Reproduction of persian literature through the art of Iranian miniature. Literature and Interarts: Critical Essays, 29-52. https:// https://publicaciones.unirioja.es/catalogo/online/interarts/pdf/02_Jamali.pdf
[5]    Rajabi, Z. and Hassanvand, M.K., (2022). An Investigation into Metaphorical Function in Human Forms of Persian Painting of Iran (Herat and Second Tabriz School). Painting Graphic Research, 4(7), 41-54. https://doi.org/10.22051/pgr.2022.40634.1138
[6]    Sandoval, C., Pirogova, E. and Lech, M., (2019). Two-stage deep learning approach to the classification of fine-art paintings. IEEE access, 7, 41770-41781. https://doi.org/10.1109/access.2019.2907986
[7]    Iman, M., Arabnia, H.R. and Rasheed, K., (2023). A review of deep transfer learning and recent advancements. Technologies, 11(2), 29-40. https://doi.org/10.3390/technologies11020040
[8]    Yao, S., Kang, Q., Zhou, M., Rawa, M.J. and Abusorrah, A., (2023). A survey of transfer learning for machinery diagnostics and prognostics. Artificial Intelligence Review, 56(4), 2859-2871. https://doi.org/10.1007/s10462-022-10230-4
[9]    Zujovic, J., Gandy, L., Friedman, S., Pardo, B., Pappas, T.N., (2009). Classifying paintings by artistic genre: An analysis of features & classifiers. In: 2009 IEEE International Workshop on Multimedia Signal Processing, 1-5. https://doi.org/10.1109/mmsp.2009.5293271
[10]    Song, H., Wang, L., & Song, C. (2025). Research on the analysis of image characteristics in fine art painting under the application of machine learning technology. Ain Shams Engineering Journal, 16(8), 103456. https://doi.org/10.1016/j.asej.2025.103456
[11]    Levy, E., David, O.E., Netanyahu, N.S., (2014). Genetic algorithms and deep learning for automatic painter classification. In: Proceedings of the 2014 Annual Conference on Genetic and Evolutionary Computation, 1143-1150. https://doi.org/10.1145/2576768.2598287
[12]    David, O.E., Netanyahu, N.S., (2016). DeepPainter: Painter classification using deep convolutional autoencoders. In: Artificial Neural Networks and Machine Learning – ICANN 2016, Part II. Springer, 20-28. https://doi.org/10.1007/978-3-319-44781-0_3
[13]    Sandoval, C., Pirogova, E., Lech, M., (2019). Two-stage deep learning approach to the classification of fine-art paintings. IEEE Access, 7, 41770-41781. https://doi.org/10.1109/access.2019.2907986
[14]    Yadav, R. K., Yadav, S. K., Yadav, M., Yadav, R., Shankar, A., & Amoon, M. (2025). Classification and feature extraction of text from hindi document for optical character recognition: RK Yadav et al. International Journal on Document Analysis and Recognition (IJDAR), 1-27. https://doi.org/10.1007/s10032-025-00552-8
[15]    Iliadis, L.A., Nikolaidis, S., Sarigiannidis, P., Wan, S. and Goudos, S.K., (2021). Artwork style recognition using vision transformers and mlp mixer. Technologies, 10(1), 12-24. https://doi.org/10.3390/technologies10010002
[16]    L Zhang, Y. (2025). Classification of Painting Image Styles Based on Hybrid Vision Transformer–ResNet50 Framework. In 2025 International Conference on Communication, Computer, and Information Technology (IC3IT) (pp. 1-7). IEEE. https://doi.org/10.1109/ic3it66137.2025.11341636
[17]    Wei, L., Li, Q., & Wang, D. (2026). Multitask Art Painting Categorization by Convolution and Transformer Hybrid Architecture. Knowledge Organization, 53(2). https://doi.org/10.31083/ko48360
[18]    Wang, H., Huang, Y., (2025). Using deep learning algorithms to identify diverse types of art designs. International Journal of Information and Communication Technology, 26(24), 15-28. https://doi.org/10.1504/ijict.2025.10071894
[19]    Zhang, Y., Lv, F., (2025). Enhancing digital art style recognition via a hybrid vision transformer and lightweight CNN with attention mechanisms. International Journal of Innovative Computing and Applications, 15(4), 236-245. https://doi.org/10.1504/ijica.2025.10073624
[20]    Wang, Z., & Song, H. (2025). A fusion model for artwork identification based on convolutional neural networks and transformers. arXiv preprint arXiv:2502.18083. https://doi.org/10.48550/arXiv.2502.18083
[21]    Parres, D., Paredes, R., (2023). Fine-tuning vision encoder-decoder transformers for handwriting text recognition on historical documents. In: Document Analysis and Recognition – ICDAR 2023. Lecture Notes in Computer Science, vol. 14190. Springer, Cham, 253-268. https://doi.org/10.1007/978-3-031-41685-9_16
[22]    Roald, M., Birkenes, M.B., Johnsen, L.B., (2024). Visual navigation of digital libraries: Retrieval and classification of images in the National Library of Norway's digitised book collection. In: Proceedings of the 5th Conference on Computational Humanities Research (CHR24), Aarhus, Denmark, December 04-06, 2024. https://doi.org/10.48550/arXiv.2410.14969%20Focus%20to%20learn%20more
[23]    Sivan, R., Pati, P.B. and Kesiman, M.W.A., (2025). Image quality determination of palm leaf heritage documents using integrated discrete cosine transform features with vision transformer. International Journal on Document Analysis and Recognition (IJDAR), 28(1), 41-57. https://doi.org/10.1007/s10032-024-00490-x
[24]    Vidal-Gorène, C., Decours-Perez, A., (2024). Detecting and deciphering damaged medieval Armenian inscriptions using YOLO and vision transformers. In: Document Analysis and Recognition – ICDAR 2024 Workshops, Lecture Notes in Computer Science, 14941, 152-168. https://doi.org/10.1007/978-3-031-70642-4_2
[25]    Schaerf, L., Popovici, C., Postma, E., (2024). Art authentication with vision transformers. Neural Computing and Applications, 36(20), 11849-11858. https://doi.org/10.1007/s00521-023-08864-8
[26]    Capriotti, S., Devoto, A., Scardapane, S., Mignardi, S., & Medeghini, L. (2025). Interpretable classification of Levantine ceramic thin sections via neural networks. Machine Learning: Science and Technology, 6(2), 025076. https://doi.org/10.1088/2632-2153/ade6c4
[27]    Touvron, H., Cord, M., Douze, M., Massa, F., Sablayrolles, A., Jégou, H., (2021). Training data-efficient image transformers & distillation through attention. In: Proceedings of the 38th International Conference on Machine Learning (ICML), 10347-10357. https://proceedings.mlr.press/v139/touvron21a
[28]    Hassani, A., Walton, S., Shah, N., Abuduweili, A., Li, J., Shi, H., (2021). Escaping the Big Data Paradigm with Compact Transformers. arXiv:2104.05704. https://doi.org/10.48550/arXiv.2104.05704%20Focus%20to%20learn%20more
[29]    Yu, L., (2026). Visual perception based deep learning transformers for classifying paintings and photographs through feature extraction. Scientific Reports (under publication). https://doi.org/10.1038/s41598-026-36298-4
[30]    Hong, S., Wu, J., Zhu, L. and Chen, W., (2024). Brain tumor classification in VIT-B/16 based on relative position encoding and residual MLP. Plos one, 19(7), p.e0298102. https://doi.org/10.1371/journal.pone.0298102
[31]    Shahi, M., Rajabi, R., Masoumzadeh, F., (2024). CNN-based classification of Persian miniature paintings from five renowned schools. Digital Applications in Archaeology and Cultural Heritage, 36, e00397. https://doi.org/10.1016/j.daach.2024.e00397
[32]    Zhang, H., Cisse, M., Dauphin, Y. N., & Lopez-Paz, D. (2017). mixup: Beyond empirical risk minimization. arXiv preprint arXiv:1710.09412. https://doi.org/10.48550/arXiv.1710.09412
[33]    Loshchilov, I., & Hutter, F. (2017). Decoupled weight decay regularization. arXiv preprint arXiv:1711.05101. https://doi.org/10.48550/arXiv.1711.05101
[34]    Tang, M., Ting, K.C. and Rashidi, N.H., (2024). DenseNet201-based waste material classification using transfer learning approach. Applied Mathematics and Computational Intelligence (AMCI), 13(2), 113-120. https://doi.org/10.58915/amci.v13i2.555
[35]    Vulli, A., Srinivasu, P.N., Sashank, M.S.K., Shafi, J., Choi, J. and Ijaz, M.F., (2022). Fine-tuned DenseNet-169 for breast cancer metastasis prediction using FastAI and 1-cycle policy. Sensors, 22(8), 2971-2988. https://doi.org/10.3390/s22082988
[36]    Khavani, S. and Marasi, M., (2025). The History of Iranian Painting as a Function of the History of Persian Literature: An Analysis of Realism in the Works of Kamal al-Din Behzad Based on the Shifts in Literary Discourse. Rahpooye Honar-Ha-Ye Tajassomi, 8(2), 45-58. https://rahpooye.soore.ac.ir/article_728299.html
[37]    Selvaraju, R.R., Cogswell, M., Das, A., Vedantam, R., Parikh, D., Batra, D., (2017). Grad-CAM: Visual Explanations from Deep Networks via Gradient-based Localization. In: Proceedings of the IEEE International Conference on Computer Vision (ICCV), 618-626. https://doi.org/10.1109/iccv.2017.74
Power, Control, and Data Processing Systems
Volume 4, Issue 1
Winter 2027
Pages 52-61

  • Receive Date 23 July 2026
  • Revise Date 11 September 2026
  • Accept Date 23 September 2026
  • First Publish Date 23 September 2026
  • Publish Date 01 March 2027

Home

AI Policy

Submit Manuscript

Contact Us