A METHOD TO IMPROVE THE ACCURACY OF TRANSFER LEARNING-BASED VGG19 FOR SKIN DISEASE CLASSIFICATION
Keywords
DOI:
https://doi.org/10.34238/tnu-jst.15003Abstract
This paper proposes an improved method based on the VGG19 deep learning model combined with transfer learning to enhance the classification of dermatological abnormalities from clinical images. In the proposed method, Contrast Limited Adaptive Histogram Equalization (CLAHE) preprocessing is applied to enhance contrast and highlight important details, thereby improving the quality of the input data for the model. Squeeze-and-Excitation (SE) blocks are integrated into the network architecture to enhance the ability to learn important features through channel weight recalibration. Additionally, a Cyclical Learning Rate strategy is used during training to optimize learning speed and improve model stability. Experiments conducted on a dataset of eight dermatological disease classes show that the proposed model achieves an accuracy of 97.85%, higher than traditional methods and some previously published related studies. These results demonstrate the potential application of deep learning in supporting the automated diagnosis of dermatological conditions in clinical practice.Downloads
References
A. M. Siddiqui, H. Abbas, M. Asim, A. A. Ateya, and H. A. Abdallah, "SGO-DRE: A Squid Game Optimization-Based Ensemble Method for Accurate and Interpretable Skin Disease Diagnosis," Computer Modeling in Engineering & Sciences, vol. 144, no. 3, pp. 3135-3168, Sep. 2025, doi: 10.32604/cmes.2025.069926.
D. Kapoor et al., "Convolutional Neural Network-based Multi-Classification of Skin Disease with Fine-Tuned ResNet50 and VGG16," Open Bioinformatics Journal, vol. 18, no. 1, July 2025, doi: 10.2174/0118750362387304250716055022.
N. Rismayanti and S. F. Azzahrah, "Classification of Skin Diseases using Decision Tree Algorithm on an Imbalanced Dataset," International Journal of Artificial Intelligence in Medical Issues (IJAIMI), vol. 2, no. 2, Nov. 2024, doi: 10.56705/ijaimi.v2i2.155.
S. S. Arjon et al., "Comprehensive Analysis of Deep Transfer Learning Models for Skin Disease Classification with Explainable AI," in 2025 International Conference on Quantum Photonics, Artificial Intelligence, and Networking (QPAIN), Rangpur, Bangladesh, 2025, doi: 10.1109/QPAIN66474.2025.11171638.
L. C. K. Pallewela, A. Senanayake, and R. Manellanga, "SkinDetective: Machine Learning Based Android Mobile Application for Automated Identification of Skin Diseases," in APIIT International Research Conference – Computing Proceedings, vol. 1, 2024, pp. 119-132. [Online]. Available: https://www.researchgate.net/publication/389401418_SkinDetective. [Accessed Feb. 14, 2026].
M. Alshahrani et al., "Analysis of dermoscopy images of multi-class for early detection of skin lesions by hybrid systems based on integrating features of CNN models," PLOS ONE, vol. 19, no. 3, Mar. 2024, doi: 10.1371/journal.pone.0298305.
J. Zhang, Y. Xie, Y. Xia, and C. Shen, "Attention Residual Learning for Skin Lesion Classification," IEEE Transactions on Medical Imaging, vol. 38, no. 9, pp. 2092-2103, Sep. 2019, doi: 10.1109/TMI.2019.2893944.
X. He, Y. Wang, S. Zhao, and C. Yao, "Deep metric attention learning for skin lesion classification in dermoscopy images," Complex & Intelligent Systems, vol. 8, no. 2, pp. 1487-1504, 2022, doi: 10.1007/s40747-021-00587-4.
J. Wu, W. Hu, Y. Wen, W. Tu, and X. Liu, "Skin Lesion Classification Using Densely Connected Convolutional Networks with Attention Residual Learning," Sensors, vol. 20, no. 24, Dec. 2020, doi: 10.3390/s20247080.
P. Tang, X. Yan, Y. Nan, X. Hu, B. H. Menze, S. Krammer, and T. Lasser, "Joint-Individual Fusion Structure with Fusion Attention Module for Multi-Modal Skin Cancer Classification," Pattern Recognition, vol. 154, 2024, doi: 10.1016/j.patcog.2024.110604.
Subir Biswas, “Skin-Disease Dataset,” 2022. [Online]. Available: https://www.kaggle.com/datasets/ subirbiswas19/skin-disease-dataset. [Accessed Feb. 14, 2026].
