ENEL645 - Data Mining & Machine Learning - Fire Detection and Classification System

Group 30

Sujesh Padhi

Abdelrahman Elnaggar

Sufiyan Bukhari

Subroto Kumer Deb Nath

Aheraz Bin Muslim Mohammad

Travis Dow

Objective

The objective of this project is to use transfer learning and use a pre-trained model named YOLOv8 for fire detection from images.

Steps to Begin

Download the fire.zip from the google drive (https://drive.google.com/file/d/1JtCrOn2jE1Pdhqru5S9oQ3KVoxECFNSL/view?usp=share_link) and place the zipped folder in your local github repository directory. Open the project.py file on your code editor for next steps.

Folder Structure

ENEL645/
├── README.md - Details of the complete project structure
│
├── project.py - main script to start training and evaluation of trained model
|
├── yolov8n.pt - base model of YOLOv8
│
├── runs/detect
│   ├── yolov8n_2/ - trained model weigth with validation results are saved here
│   |── validation/ - testing results are saved here
|   └── all the directories contain results from various epoch runs

Artefacts

GitHub Repository – https://github.com/ttdow/ENEL645

Fire Dataset Location – https://drive.google.com/file/d/1JtCrOn2jE1Pdhqru5S9oQ3KVoxECFNSL/view?usp=share_link

Presentation video – https://youtu.be/etBiSHpHKi0

Fire detection video – https://www.youtube.com/watch?v=B8NxlKMXVjU