Towards Efficient Blood Cell Detection: A Parameter-Reduced YOLOv8n Model with GhostConv and C3Ghost

Published: 5 November 2025| Version 1 | DOI: 10.17632/jrwz8dpm9k.1
Contributors:
Jing Yang, Zhenqing Li

Description

The dataset and source code used in this study support blood cell analysis based on an enhanced YOLOv8n model. The dataset is stored in the designated directory and includes microscopic blood cell images from the publicly available BCCD database (accessible at https://datasetninja.com/bccd), which are annotated in YOLO format to enable supervised training. The corresponding code for model training and prediction is also provided.

Files

Categories

Deep Learning

Licence