aakinlalu/Imageobject-Classification-Detection

Notebooks to fine-tune Yolov5 and Yolov8 and export model for JavaScript deployment

Custom the Weights for Object Detection using yolov5 or yolov8

This includes how to train labelled images from ROBOFLOW Platform, evaluate the results and export to onnx or tensorflowjs

1. Clone the repo and cd into

cd object-classifier-tng

2. Set up Python Environemnt

Assuming using Python3.8 and above. This is setup with Python 3.10.9

python3 -m venv object-classifier-env

Activate it

source object-classifier-env/bin/activate

3. Install dependencies

pip install -r requirements-dev.txt
pip install -r requirements.txt 

4. For Yolov5 (jupyter)

cd notebook-yolov5 && jupyter lab 

Or for Yolov5 (Vscode)

cd notebook-yolov5 && code .

5. For Yolov8 (jupyter)

cd notebook-yolov8 && jupyter lab 

Or for Yolov8 (Vscode)

cd notebook-yolov8 && code .

6. Use JavaScript API to load and make prediction

6a. Use ml5.js and p5.js -> click here

6n. Use ml5.js and p5.js -> click here