Matching_and_Damage / README.md
RehamAAhmed's picture
Upload 4 files
b8760c9 verified
|
Raw
History Blame Contribute Delete
1.59 kB
---
title: Property Damage Inspector & Matcher
emoji: 🏠
colorFrom: blue
colorTo: green
sdk: docker
app_port: 7860
pinned: false
---
# 🏠 AI Property Damage Inspector & Inventory Matcher
An advanced computer vision pipeline combining **DINOv2**, **Grounded-SAM**, and **MLLMs (Gemini, Llama Vision, GPT-4o)** to automate inventory matching and damage assessment between "Before" and "After" states.
## 🌟 Key Features
- **Directory-wide Batch Comparison:** Matches items between two folders automatically using a DINOv2 cls, patch matching, and SIFT+RANSAC cascade.
- **AI-Powered Damage Detection:** Inspects matched pairs for deep scratches, breakages, or tears using multi-API providers (Gemini, OpenRouter, OpenAI, Groq).
- **Damage Segmentation (Grounded-SAM):** Automatically generates precise segmentation masks where damage is located using Grounding DINO + SAM.
- **Premium Bilingual Reports:** Exports a consolidated, interactive, self-contained dark-mode HTML report (English & Arabic) embedding all base64-encoded images.
## πŸš€ Deployment on Hugging Face Spaces
This repository is configured to deploy directly to Hugging Face Spaces using **Docker**.
### How to Deploy:
1. Create a new Space on [Hugging Face](https://huggingface.co/new-space).
2. Select **Docker** as the SDK.
3. Upload all the files from this directory.
4. Add your API keys (e.g., `OPENROUTER_API_KEY`, `GEMINI_API_KEY`) under **Settings > Variables and secrets** in your Hugging Face Space if you want to preload them.
5. Hugging Face will build the container and start the web UI automatically.