Resolution-Aware Malaria Screening: Do Super-Resolved RBC Images Improve CNNs and Vision Transformers?

• Igor Oliveira UFOP
• Arthur Negrão UFOP
• Ederson N. F. G. Júnior UFOP
• Guilherme Silva UFOP
• Matheus Vieira UFOP
• Pedro Silva UFOP

DOI: https://doi.org/10.5753/sbcas.2026.21321

Abstract

Automated malaria screening from microscopy images can improve diagnostic scalability in resource-limited settings, but detecting infected red blood cells (RBCs) is challenging due to subtle morphological cues and resolution constraints. We study resolution-aware malaria classification on the NIH Malaria Dataset using EfficientNet-B0 and ViTs, evaluating transfer learning, optimization strategies, and ESRGAN-based super-resolution preprocessing. Results show that cosine decay and transfer learning are critical for robust performance, enabling EfficientNet-B0 to reach 96% accuracy and recall, while ESRGAN with ViT achieves 97% accuracy and 98% recall, matching SOTA results while reducing false negatives.

References

Çinar, A. and Yildirim, M. (2020). Classification of malaria cell images with deep learning architectures. Ingénierie des Systèmes d’Information, 25(1):35.

Dosovitskiy, A., Beyer, L., Kolesnikov, A., Weissenborn, D., Zhai, X., Unterthiner, T., Dehghani, M., Minderer, M., Heigold, G., Gelly, S., et al. (2020). An image is worth 16x16 words: Transformers for image recognition at scale. arXiv preprint arXiv:2010.11929.

Harahap, M., Jefferson, J., Barti, S., Samosir, S., and Turnip, C. A. (2021). Implementation of convolutional neural network in the classification of red blood cells have affected of malaria. Sinkron: jurnal dan penelitian teknik informatika, 5(2):199–207.

Hugging Face (2020). vit-base-patch16-224-in21k. Accessed on: March, 2026.

Minarno, A. E., Aripa, L., Azhar, Y., and Munarko, Y. (2023). Classification of malaria cell image using inception-v3 architecture. JOIV: International Journal on Informatics Visualization, 7(2):273–278.

Rajaraman, S., Antani, S. K., Poostchi, M., Silamut, K., Hossain, M. A., Maude, R. J., Jaeger, S., and Thoma, G. R. (2018). Pre-trained convolutional neural networks as feature extractors toward improved malaria parasite detection in thin blood smear images. PeerJ, 6:e4568.

Reddy, A. S. B. and Juliet, D. S. (2019). Transfer learning with resnet-50 for malaria cell-image classification. In 2019 International conference on communication and signal processing (ICCSP), pages 0945–0949. IEEE.

Secretaria de Estado da Saúde do Pará (2023). O que é a malária? [link]. Accessed on: January, 2026.

Shekar, G., Revathy, S., and Goud, E. K. (2020). Malaria detection using deep learning. In 2020 4th international conference on trends in electronics and informatics (ICOEI)(48184), pages 746–750. IEEE.

Tan, M. and Le, Q. (2019). Efficientnet: Rethinking model scaling for convolutional neural networks. In International conference on machine learning, pages 6105–6114. PMLR.

Wang, X., Yu, K., Wu, S., Gu, J., Liu, Y., Dong, C., Qiao, Y., and Loy, C. C. (2018). Esrgan: Enhanced super-resolution generative adversarial networks. In Proceedings of the European Conference on Computer Vision (ECCV) Workshops, pages 0–0.

Waseem Sabir, M., Farhan, M., Almalki, N. S., Alnfiai, M. M., and Sampedro, G. A. (2023). Fibrovit—vision transformer-based framework for detection and classification of pulmonary fibrosis from chest ct images. Frontiers in medicine, 10:1282200.