Object Detecion with YOLO

This project build a YOLO system that can detect fruit that is durian, apples, dragon fruits, banana, and oranges

Overview

YOLO is a Deeplearning Architecture that can detect object using IoU post-process. However, later YOLO like YOLOv10, this post-processing mechanism is cut. This project uses pre-trained weight from YOLO8n and finetune for fruit detection task.

DATA

Dataset is collected by download random fruit online.

Train Model

in notebook

result

precision: 64.6 \ recall: 42.76

mAP50: 44.06 \ mAP50-95: 18.48

fitness: 21.04

Using

install requirements pip install -r requirements.txt

you can further train this model in notebook

you can also run the main script for UI inference section.