A CNN-Based Visual Framework for Peking Opera Facial Mask Recognition with Multi-Feature Cultural Representation
Main Article Content
Keywords
Peking Opera masks, cultural visual recognition, convolutional neural networks, transfer learning, feature bias
Abstract
Peking Opera facial masks constitute a highly symbolic visual system in traditional Chinese performing arts, where color and structural patterns encode rich cultural meanings such as character identity and moral traits. In recent years, advances in compute r vision and deep learning have enabled increasing interest in applying automated methods to the recognition and analysis of such culturally significant visual forms. This paper suggests the usage of a convolutional neural network (CNN)-based system to identify Peking Opera face masks through a combination of color, texture, structural symmetry, features. The framework attempts to overcome the shortcomings of generic models of deep learning in the ability to capture culturally based visual semantics. The model is created on a ResNet backbone and the transfer learning approach, which is aimed at small-sample cultural data. Experimental observations of related works indicate that the use of domain specific representations enhances classification accuracy and minimizes color bias that is evident in CNN models. The results show that CNNs are useful in visual pattern recognition, but they have a poor capability of deciphering cultural meaning. The paper emphasizes the need to integrate computing techniques with cultural priors to augment cultural visual recognition tasks in their interpretability and strength.
References
- [1] Wang, D., Kang, J., Qin, S. F., & Birringer, J. (2016). Cultural-based visual expression: emotional analysis of human face via Peking Opera Painted Faces (POPF). Multimedia Tools and Applications, 75(19), 11865-11891.
- [2] Turay, T., & Vladimirova, T. (2022). Toward performing image classification and object detection with convolutional neural networks in autonomous driving systems: A survey. IEEE access, 10, 14076-14119.
- [3] Khan, M. S., & Imran, A. (2024). The Art of Seeing: A Computer Vision Journey into Object Detection. Adv Mach Lear Art Inte, 5(2), 1.
- [4] Nie, X., & Zhao, L. N. (2026). Research on Virtual Costume Design for Traditional Cultural Symbols: An Intangible Heritage Display Method Based on Computer Vision Technology. International Journal of Computer Information Systems and Industrial Management Applications, 18, 12-12.
- [5] Pan, H., Pan, S., & Pang, B. (2025). The Application of Computer Vision and CAD Technology in the Interactive Platform for Calligraphy Art. Computer-Aided Design & Applications, 22.
- [6] Du, X., & Cai, Y. (2024). Design of Chinese painting style classification model based on multi-layer aggregation CNN. PeerJ Computer Science, 10, e2303.
- [7] He, K., Zhang, X., Ren, S., & Sun, J. (2016). Deep residual learning for image recognition. In Proceedings of the IEEE conference on computer vision and pattern recognition (pp. 770-778).
- [8] Geirhos, R., Rubisch, P., Michaelis, C., Bethge, M., Wichmann, F. A., & Brendel, W. (2019). ImageNet- trained CNNs are biased towards texture; increasing shape bias improves accuracy and robustness. arXiv preprint arXiv:1811.12231.
- [9] Liao, J., Guo, L., Jiang, L., Yu, C., Liang, W., Li, K., & Pop, F. (2025). A machine learning-based feature extraction method for image classification using ResNet architecture. Digital Signal Processing, 160, 105036.
- [10] Zhong, G., Wang, L. N., Ling, X., & Dong, J. (2016). An overview on data representation learning: From traditional feature learning to recent deep learning. The Journal of Finance and Data Science, 2(4), 265 - 278.
- [11] Asadi, B., & Jiang, H. (2020). On approximation capabilities of ReLU activation and softmax output layer in neural networks. arXiv preprint arXiv:2002.04060.
- [12] Jeevan, P., & Sethi, A. (2024). Which backbone to use: a resource-efficient domain specific comparison for computer vision. arXiv preprint arXiv:2406.05612.
- [13] Milani, F., & Fraternali, P. (2021). A dataset and a convolutional model for iconography classification in paintings. Journal on Computing and Cultural Heritage (JOCCH), 14(4), 1-18.
- [14] Zhang, K., & Shi, L. (2024, September). Research on Style Transfer Method from Human Facial Images to Peking Opera Facial Images. In 2024 5th International Conference on Intelligent Computing and Human-Computer Interaction (ICHCI) (pp. 185-190). IEEE.
- [15] Qi, L., Wang, Y., Chen, J., Liao, M., & Zhang, J. (2021). Culture under complex perspective: a classification for traditional chinese cultural elements based on nlp and complex networks. Complexity, 2021(1), 6693753.
- [16] Alam, M., Alourani, A., Ali, A., & Ahamad, F. (2025). Symmetry-Aware Feature Representations and Model Optimization for Interpretable Machine Learning. Symmetry, 17(11), 1821.
Article Sidebar
This work is licensed under a Creative Commons Attribution 4.0 International License.