Model Card for YOLOv8m - Harmful Smoke Detection in Indian Vehicles

This model detects harmful smoke emissions from various types of Indian vehicles. It is built on the YOLOv8m architecture and trained on a custom dataset.

Model Details

Model Description

The YOLOv8m model is designed to identify harmful smoke emissions from different types of Indian vehicles, including Auto-Rickshaws, Bikes, Buses, Cars, Heavy Commercial Vehicles (HCV), Light Commercial Vehicles (LCV), and Totos. The custom dataset used for training was gathered from the internet, focusing specifically on smoke detection in these vehicle categories.

Developed by: HarshitJoshi
Model type: Object Detection (YOLOv8m)

Results

The model's performance is evaluated based on several metrics: Precision (P), Recall (R), Mean Average Precision at IoU=0.5 (mAP50), and Mean Average Precision at IoU range 0.5:0.95 (mAP50-95).

Summary

Class	Images	Instances	Box(P)	Box(R)	mAP50	mAP50-95
All	952	1422	0.731	0.678	0.749	0.441
Auto-Rickshaw	119	138	0.679	0.812	0.83	0.496
Bike	195	278	0.89	0.643	0.809	0.412
Bus	202	307	0.881	0.435	0.667	0.422
Car	177	300	0.832	0.797	0.843	0.486
HCV	146	195	0.89	0.497	0.72	0.377
LCV	40	42	0.3	0.762	0.55	0.292
Toto	137	162	0.642	0.796	0.826	0.598

Model Architecture and Objective

The model is based on the YOLOv8m architecture, known for its speed and accuracy in object detection tasks. The primary objective of this model is to detect harmful smoke emissions from various types of vehicles commonly found in India.

Base Architecture: YOLOv8m
Dataset: Custom dataset gathered from the internet
Training Objective: Detect harmful smoke emissions from Indian vehicles