AutoClaim AI

Project Overview

AutoClaim AI is a full-stack computer vision application I developed to automatically detect vehicle damage from images and estimate insurance claim amounts. The goal of the project is to simulate a car damage assessment tool that streamlines the insurance process and reduces the need for manual inspection.

Key Features:

• Upload or capture a photo of a damaged car

• Automatically detect damage using a YOLOv8 object detection model

• Classify the type and severity of damage (e.g., scratch, dent, crack)

• Estimate insurance claim costs using rule-based logic and regression modeling

• Display annotated results with bounding boxes and a claim summary

Tools & Technologies Used:

• Computer Vision: YOLOv8, OpenCV

• Machine Learning: TensorFlow / PyTorch, Scikit-learn

• API Backend: Python, FastAPI

• Visualization: Matplotlib, OpenCV

What I Learned:
This project deepened my practical knowledge of object detection pipelines, severity classification, and deploying computer vision models in a full application workflow.

Data Sources

This project uses publicly available datasets from Kaggle.

1. COCO Car Damage Detection Dataset

Annotated car images with bounding boxes around visible damage like dents, scratches, and cracks.
Used for training the damage localization model.
🔗 View Dataset on Kaggle

2. Car Damage Severity Dataset

Image dataset labeled by severity levels: Minor, Moderate, and Severe.
Used to train a severity classification model.
🔗 View Dataset on Kaggle

3. Manually Created Cost Mapping Table

A custom table mapping damage type and severity to estimated repair costs.

Code

coming soon

Note to Hiring Managers:
Thank you for taking the time to explore my work. AutoClaim AI is one example of how I apply data science and computer vision to solve problems with impact. I'm passionate about building intelligent systems that are practically useful. If you're looking for someone who is looking to hire a computer vision passionate, I’d love to connect and explore opportunities with your team.

How It Works

Learn about the AI algorithms used for predictions.

Technical Breakdown

Detailed insights into the project's technical architecture.

Source Code

Access project files and source code for development.

→

→

→