Comparative Analysis of Convolutional Neural Network Architectures: The Impact of Hair and Noise Removal on Algorithm Learning

• Matheus Rodrigues de Oliveira Solon IFCE
• Vivian Kailany Marques de Queiroz IFCE
• Roney Nogueira Sousa IFCE

DOI: https://doi.org/10.5753/enucompi.2025.9621

Abstract

This work presents a comparative analysis of different Convolutional Neural Network architectures — ResNet50V2, InceptionV3, EfficientNetB7, and MobileNetV2 — aiming to evaluate the impact of dermoscopic image preprocessing, particularly hair and noise removal, on classification performance. Four experimental scenarios were considered, with and without the use of transfer learning and data augmentation. Results showed that ResNet50V2, when combined with preprocessing and deep learning strategies, achieved the best performance. The study concludes that proper image treatment is crucial to improve accuracy and the discriminative capability of CNN models.

References

Abadi, M., Barham, P., Chen, J., Chen, Z., Davis, A., Dean, J., Devin, M., Ghemawat, S., Irving, G., Isard, M., et al. (2016). Tensorflow: a system for large-scale machine learning. In 12th USENIX symposium on operating systems design and implementation (OSDI 16), pages 265–283, Savannah, GA. USENIX Association.

Bradski, G. (2000). The OpenCV Library. Dr. Dobb’s Journal of Software Tools.

Carvalho, M. B. F. d. (2020). Reconhecimento automático de fonemas via rna profunda. Biblioteca Digital de Teses e Dissertações UFMA.

Chollet, F. et al. (2015). Keras. [link].

da Saúde, M. (2024). Câncer de pele. [link]. [online; acesso em 09 de março de 2025].

da Saúde, O. M. (2025). Cancer. [link]. [online; acesso em 09 de março de 2025].

DiMatos, D. C. and Al., E. (2009). Melanoma cutâneo no brasil. Arquivos Catarinenses de Medicina, 38(Suplemento 01):14.

Esteva, A., Kuprel, B., et al. (2017). Dermatologist-level classification of skin cancer with deep neural networks. Nature, 542(7639):115–118.

G. Argenziano, H. P. S. (2001). Dermoscopy of pigmented skin lesions–a valuable tool for early diagnosis of melanoma. The Lancet Oncology, 2(7):443–449.

Goodfellow, I., Bengio, Y., and Courville, A. (2016). Deep learning. MIT press.

He, K., Zhang, X., Ren, S., and Sun, J. (2016). Deep residual learning for image recognition. Proceedings of the IEEE conference on computer vision and pattern recognition, pages 770–778.

Howard, A. G., Zhu, M., Chen, B., Kalenichenko, D., Wang, W., Weyand, T., Andreetto, M., and Adam, H. (2017). Mobilenets: Efficient convolutional neural networks for mobile vision applications. arXiv preprint arXiv:1704.04861.

Hunter, J. D. (2007). Matplotlib: A 2d graphics environment. Computing in Science & Engineering, 9(3):90–95.

LeCun, Y., Bottou, L., Bengio, Y., and Haffner, P. (1998). Gradient-based learning applied to document recognition. Proceedings of the IEEE, 86(11):2278–2324.

Sandler, M., Howard, A., Zhu, M., Zhmoginov, A., and Chen, L.-C. (2018). Mobile-netv2: Inverted residuals and linear bottlenecks. Proceedings of the IEEE conference on computer vision and pattern recognition, pages 4510–4520.

Souza, V., Silva, L., Santos, A., and Araújo, L. (2020). Análise comparativa de redes neurais convolucionais no reconhecimento de cenas. Anais do Computer on the Beach, 11(1):419–426.

Szegedy, C., Liu, W., Jia, Y., Sermanet, P., Reed, S., Anguelov, D., Erhan, D., Vanhoucke, V., and Rabinovich, A. (2015). Going deeper with convolutions. Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition (CVPR), pages 1–9.

Tan, M. and Le, Q. V. (2019). Efficientnet: Rethinking model scaling for convolutional neural networks. CoRR, abs/1905.11946.

Van Rossum, G. and Drake, F. L. (2009). Python 3 Reference Manual. CreateSpace, Scotts Valley, CA.

Vieira, E. Q. (2022). Comparação entre diferentes modelos de redes neurais convolucionais para classificação de melanoma. Universidade de Brasília.

Zhong, Z., Jin, L., and Xie, Z. (2015). High performance offline handwritten chinese character recognition using googlenet and directional feature maps. In 2015 13th International Conference on Document Analysis and Recognition (ICDAR), pages 846–850. IEEE.