README.md

---
tags:
- Tyre Quality
- Tyre Classification
- Image Classification
- Machine Learning
- Pytorch
- Deep Learning
- Computer Vision
- Prediction
- yolov8
- yolo
- TyreInspection
- QualityControl
- DefectDetection
- AutomotiveAI
- SafetyStandards
- IndustrialAI
- AIQualityAssessment
- PredictiveMaintenance
- AIModel
library_name: ultralytics
library_version: 8.0.43
language:
- en
pipeline_tag: image-classification
model-index:
- name: foduucom/Tyre-Quality-Classification
  results:
  - task:
      type: image-classification
    metrics:
    - type: accuracy
      value: 0.835
      name: Top1_acc
---

<div align="center">
  <img width="640" alt="foduucom/Tyre-Quality-Classification-AI" src="https://huggingface.co./foduucom/Tyre-Quality-Classification-AI/resolve/main/Tyre-Quality-Classification-AI%20(3).jpg">
</div>

### Welcome to the repository of our state-of-the-art image classification model, uniquely fine-tuned on the robust architecture of YOLOv8s, tailored to distinguish between defective and good tyres with unprecedented accuracy.

# Model Details

## Model Description

Our model leverages the cutting-edge capabilities of YOLOv8s, renowned for its speed and precision in object detection, which has been meticulously fine-tuned for the specific domain of tyre quality assessment. This model emerges as an indispensable tool for automating quality control in tyre manufacturing, ensuring that every tyre meets the highest standards of safety and performance.

- **Developed by:** FODUU AI
- **Model type:** Image Classification
- **Task:**  Classifies tyres with a high degree of accuracy


### Supported Labels

```
['Good_Tyre','Defective_Tyre']
```


## Key Features:
- High Precision Classification: Classifies tyres with a high degree of accuracy, reducing the margin of error significantly compared to traditional methods.
- Rapid Assessment: Optimized for quick image processing, allowing for real-time quality control on the production line.
- Robust Training: Trained on a diverse dataset of tyre images, capturing a wide range of defects to ensure reliability in various operational environments.
- Easy Integration: Designed to seamlessly fit into existing manufacturing systems, facilitating a smooth transition from manual to automated quality assessment.


# Model Training Details

## Dataset Description
Our model was trained on a comprehensive and proprietary dataset consisting of several images, each meticulously labeled as 'defective' or 'good'. The dataset encompasses a variety of tyre types and conditions to ensure broad applicability and robustness. It includes:

Images of good tyres: High-quality images captured under various lighting and background conditions to simulate real-world scenarios.
Images of defective tyres: A diverse range of defects such as tread wear, sidewall bulges, and punctures, among others, were included to train the model effectively.

## Preprocessing Steps
Prior to training, we applied several preprocessing steps to standardize the dataset and enhance model performance:


Augmentation: To increase the robustness of our model, we applied random rotations, flips, and color adjustments to simulate a variety of conditions.
Normalization: Pixel values were normalized to aid in the model's faster convergence during training.

## Training Procedure
The model was trained using the following procedure:

Base Model: We used the YOLOv8s-cls pre-trained on the COCO dataset as our starting point.
Fine-Tuning: The model was then fine-tuned on our specific tyre image dataset, indicating optimal learning.
Batch Size: A particular batch size was employed to make the best use of our GPU memory resources.
Optimizer: SGD optimizer was used for its fast convergence properties.


## How to Use This Model
This model is hosted on Hugging Face, making it incredibly easy to integrate and deploy. You can directly use our pre-trained model for classifying your tyre images by following the instructions in our usage documentation.

### Getting Started

```bash
pip install ultralyticsplus==0.0.28 ultralytics==8.0.43
```
- Load model and perform real-time prediction:

```python
from ultralyticsplus import YOLO, render_result
import torch


# load model
model = YOLO('foduucom/Tyre-Quality-Classification-AI')

# set model parameters
model.overrides['conf'] = 0.25  # NMS confidence threshold
model.overrides['iou'] = 0.45  # NMS IoU threshold
model.overrides['agnostic_nms'] = False  # NMS class-agnostic
model.overrides['max_det'] = 1000  # maximum number of detections per image

# set image
image = '/content/90-100-10-ceat-500x500.jpeg'

# perform inference
results = model(image)


# observe results
top_class_index = torch.argmax(results[0].probs).item()
Class = model.names[top_class_index]
print(Class)

```



## Performance Metrics
Our model achieved the following performance metrix on the test set:

- Accuracy: 83.5%
This metrix demonstrate the model's reliability in correctly classifying tyres as defective or good.
Hardware and Software
Training Environment: The model was trained on a machine with an NVIDIA RTX 3060 GPU, 32GB RAM, and an AMD Ryzen 5 processor.


### Support and Contributions
We are committed to the continuous improvement of our model and welcome contributions from the community. If you encounter any issues or would like to contribute, please refer to our contribution guidelines.

### Stay Updated
Don't forget to ⭐ this repository if you find this model useful! Follow this repo to stay updated with the latest enhancements and releases.

Start harnessing the power of AI for tyre defect detection today – Download and integrate our fine-tuned YOLOv8s model now!

### Model Card Contact
For inquiries and contributions, please contact us at [email protected].
```bibtex
@ModelCard{
  author    = {Nehul Agrawal, Rishabh Karnawat and
               Rahul parihar},
  title     = {Fine-Tuned YOLOv8s for Tyre Quality Classification},
  year      = {2023}
}
```