Copyright © 2026 Authors retain the copyright of this article. This article is an open access article distributed under the Creative Commons Attribution License which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

@article{209622,
        author = {Dr. M. K. Jayanthi Kannan and Gowri Nair},
        title = {MineDL: Class-Rarity Adaptive Ensemble Learning for Blood Cell Detection Using YOLO and Faster R-CNN},
        journal = {International Journal of Innovative Research in Technology},
        year = {2026},
        volume = {13},
        number = {5},
        pages = {3771-3780},
        issn = {2349-6002},
        url = {https://ijirt.org/article?manuscript=209622},
        abstract = {Automated blood cells detection helps with the consistent analysis of blood smear images. Manual examination is quite time-consuming. In this project, we propose a deep learning ensemble framework that combines three object detection models for detecting mainly three cells – White Blood Cells (WBC), Red Blood Cells (RBC), and Platelets. The images used here are from the TXL-PBC dataset of 1260 images that are already annotated. The models in the ensemble are YOLOv8n, YOLOv11n, and Faster R-CNN. The predictions of these models were combined using Weighted Boxes Fusion (WBF) as a baseline and a Class Rarity Adaptive Weighted Boxes Fusion (CRA-WBF). The CRA-WBFcomputes fusion weights per model and per class. The models have achieved validation mAP50 of 98.85%, 98.51%, and 96.34%. On the test set, WBF achieved an overall mAP50 of 97.409% and CRA-WBF 97.411%.},
        keywords = {},
        month = {October},
        }