Fruit Detection Model - Machine Learning Project

Machine Learning • YOLO • PyTorch

yolo 1
yolo 3
yolo 4
yolo 4
yolo 5

Project Overview

I needed a detector trained on my own fruit images. I labeled 100 photos in Label Studio, trained a YOLO model for 60 epochs on Google Colab, and deployed it with PyTorch on my RTX 4060. It detected fruit live through a webcam.

Key Features

What I Learned

I learned to build a dataset by varying lighting and backgrounds, then label each image consistently in Label Studio. Training in Colab and running the model locally taught me to read loss graphs and manage the Python and PyTorch setup needed for live webcam detection.