Detecção da psoríase utilizando visão computacional: Uma abordagem comparativa entre CNNS e Vision Transformers
| dc.contributor.advisor | Silva, Fábio Santos da | |
| dc.contributor.author | Medeiros, Natanael Lucena de | |
| dc.contributor.author-lattes | http://lattes.cnpq.br/2949133572831254 | |
| dc.contributor.referee1 | Silva, Fábio Santos da | |
| dc.contributor.referee2 | Figueiredo, Maurício Serodio | |
| dc.contributor.referee3 | Melo, Tiago Eugenio de | |
| dc.date.accessioned | 2026-08-07T18:40:09Z | |
| dc.date.issued | 2025-04-10 | |
| dc.description.abstract | This work performs a comparative analysis between Convolutional Neural Networks (CNNs) and Vision Transformers (ViT) models on the task of multiclassifying skin images, detecting psoriasis among other images of skin diseases visually similar to it, such as dermatitis, lichen planus and pityriasis rosea. The aim of the experiment is to identify the most suitable model for detecting psoriasis, so that it can be used as a diagnostic tool for the disease. During implementation, each model is pre-trained from ImageNet and adapted to the collected image set. During the comparative analysis, it is observed that both CNN and ViT architectures present excellent prediction metrics in the experiment, but ViTs stand out by achieving slightly better performance with significantly smaller models. Finally, DaViT-B (Dual Attention Vision Transformer - Base) stood out from the rest and was therefore selected as the most suitable model for detecting psoriasis, with an f1-score of 96.4%. | |
| dc.description.resumo | Esse trabalho apresenta uma análise comparativa entre modelos de Rede Neurais Convolucionais (CNNs) e de Vision Transformers (ViT) no que diz respeito `a tarefa de multiclassificaçao de imagens de pele com psoríase e similares, como, por exemplo, dermatite, líquen plano e pitiríase rosada. O objetivo do experimento ´e identificar o modelo mais indicado para a detecção de psoríase, como ferramenta de diagnóstico da doença. Na implementação, cada modelo ´e pré-treinado a partir do ImageNet e adaptado para o conjunto de imagens coletado. Durante a análise comparativa, observou-se que ambas as arquiteturas, CNN e ViT, apresentam ´ótimas métricas de predição nos experimentos, mas os ViTs se destacam por atingir um desempenho ligeiramente superior com modelos significativamente menores. Por fim, o DaViT-B (Dual Attention Vision Transformer - Base) se destacou entre todos, sendo, portanto, selecionado como o modelo mais indicado para a detecção da psoríase, apresentando um f1-score de 96,4%. | |
| dc.identifier.citation | MEDEIROS, Natanael Lucena de. Detecção da psoríase utilizando visão computacional: Uma abordagem comparativa entre CNNS e Vision Transformers. 2024. Digital. TCC (Graduação em Engenharia de Computação) - Universidade do Estado do Amazonas, Manaus, 2024. | |
| dc.identifier.uri | https://ri.uea.edu.br/handle/riuea/8562 | |
| dc.publisher | Universidade do Estado do Amazonas | |
| dc.publisher.initials | UEA | |
| dc.relation.references | SILVA, S. R. e. Modelagem de oxigênio dissolvido utilizando redes neurais artificiais. Dissertação (Mestrado) — Instituto Federal de Goiás, 2014. SILVA, T. Um estudo comparativo entre algoritmos de aprendizagem de máquina supervisionados para predição de solução de reclamações no procon. Centro Universitário Christus-Unichristus, 2021. SILVERBERG, J. I. Public health burden and epidemiology of atopic dermatitis. Dermatologic clinics, Elsevier, v. 35, n. 3, p. 283–289, 2017. SIMONYAN, K. Very deep convolutional networks for large-scale image recognition. arXiv preprint arXiv:1409.1556, 2014. SMITH, L. N. Cyclical learning rates for training neural networks. In: IEEE. 2017 IEEE winter conference on applications of computer vision (WACV). [S.l.], 2017. p. 464–472. SOBREIRO, V. A. et al. Uma estimação do valor da commodity de açúcar utilizando redes neurais artificiais. Pesquisa & Desenvolvimento em Engenharia de Produção, v. 7, p. 36–52, 2008. SOUZA, E. et al. Análise histológica comparativa entre epitélios saudáveis e doentes com psoríase e líquen plano. Rev. Ciênc. Saúde Nova Esperança.[Internet], p. 1–15, 2016. SOUZE, G. G. B. Deep Learning para a Detecção e Classificação de Pneumonia por Radiografias do Tórax. Dissertação (Mestrado) — Universidade Federal do Maranhão, 2018. STEFANINI, M. et al. From show to tell: A survey on deep learning-based image captioning. IEEE transactions on pattern analysis and machine intelligence, IEEE, v. 45, n. 1, p. 539–559, 2022. STEWART JULIA BATTAGLINI-SABETTA, L. M. A. F. Hypocalcemia-induced pustular psoriasis of von zumbusch. Annals of Internal Medicine, v. 100, n. 5, p. 677–680, 1984. PMID: 6546844. Disponível em: <https://www.acpjournals.org/doi/abs/10.7326/0003-4819-100-5-677>. SZEGEDY, C. et al. Inception-v4, Inception-ResNet and the Impact of Residual Connections on Learning. 2016. SZEGEDY, C. et al. Rethinking the inception architecture for computer vision. In: Proceedings of the IEEE conference on computer vision and pattern recognition. [S.l.: s.n.], 2016. p. 2818–2826. TAN, M.; LE, Q. Efficientnet: Rethinking model scaling for convolutional neural networks. In: PMLR. International conference on machine learning. [S.l.], 2019. p. 6105–6114. TAN, M.; LE, Q. Efficientnetv2: Smaller models and faster training. In: PMLR. International conference on machine learning. [S.l.], 2021. p. 10096–10106. TEAM, G. et al. Gemini 1.5: Unlocking multimodal understanding across millions of tokens of context. arXiv preprint arXiv:2403.05530, 2024. TSCHANDL, P.; ROSENDAHL, C.; KITTLER, H. The ham10000 dataset, a large collection of multi-source dermatoscopic images of common pigmented skin lesions. Scientific data, Nature Publishing Group, v. 5, n. 1, p. 1–9, 2018. TU, Z. et al. Maxvit: Multi-axis vision transformer. In: SPRINGER. European conference on computer vision. [S.l.], 2022. p. 459–479. VASWANI, A. Attention is all you need. Advances in Neural Information Processing Systems, 2017. WASKOM, M. L. seaborn: statistical data visualization. Journal of Open Source Software, The Open Journal, v. 6, n. 60, p. 3021, 2021. Dispon´ıvel em: <https://doi.org/10.21105/joss.03021>. WASSERMAN, P. D.; SCHWARTZ, T. Neural networks. ii. what are they and why is everybody so interested in them now? IEEE expert, IEEE, v. 3, n. 1, p. 10–15, 1988. WEB docente de Dermatologia. 2002. <http://dermatoweb.udl.es/>. Acesso em: 22/04/2024. WEI, M. et al. A skin disease classification model based on densenet and convnext fusion. Electronics, MDPI, v. 12, n. 2, p. 438, 2023. WIGHTMAN, R. PyTorch Image Models. [S.l.]: GitHub, 2019. <https://github.com/ rwightman/pytorch-image-models>. WISSTEIN, E. W. Convolution. 2024. <https://mathworld.wolfram.com/Convolution.html>.MathWorld–A Wolfram Web Resource. Acesso em: 28/04/2024. WOOD, T. What is the Softmax Function? 2018. <https://deepai.org/machine-learning-glossary-and-terms/softmax-layer>. Acesso em: 23 de nov. de 2024. WU, Z. et al. Studies on different cnn algorithms for face skin disease classification based on clinical images. IEEE Access, v. 7, p. 66505–66511, 2019. XIN ZIHAN ZHOU, Y. X. Y. Scene separation & data selection: Temporal segmentation algorithm for real-time video stream analysis. CEUR Workshop Proceedings, 07 2023. YEUNG, H. et al. Psoriasis severity and the prevalence of major medical comorbidity: A population-based study. JAMA Dermatology, American Medical Association (AMA), v. 149, n. 10, p. 1173, out. 2013. ISSN 2168-6068. Dispon´ıvel em: <http://dx.doi.org/10.1001/jamadermatol.2013.5015>. ZHAO, S. et al. Smart identification of psoriasis by images using convolutional neural networks: a case study in china. Journal of the European Academy of Dermatology and Venereology, Wiley Online Library, v. 34, n. 3, p. 518–524, 2020. ZHOU, R.; KHEMMARAT, S.; GAO, L. The impact of youtube recommendation system on video views. In: . [S.l.: s.n.], 2010. p. 404–410. | |
| dc.subject | Inteligência Artificial | |
| dc.subject | Redes Neurais Convolucionais | |
| dc.subject | Transformers | |
| dc.subject | Visão Computacional | |
| dc.subject | Psoríase. | |
| dc.title | Detecção da psoríase utilizando visão computacional: Uma abordagem comparativa entre CNNS e Vision Transformers | |
| dc.type | Trabalho de Conclusão de Curso |
Arquivos
Pacote original
1 - 1 de 1
Carregando...
- Nome:
- Detecção_psoríase_visão_computacional.pdf
- Tamanho:
- 5.35 MB
- Formato:
- Adobe Portable Document Format