---
license: mit
tags:
- anomaly-detection
- quality-control
- computer-vision
- efficientad
---

# EfficientAD - Bottle Inspection Model

Production-ready anomaly detection model for bottle quality inspection.

## Model Details
- **Dataset**: MVTec AD - Bottle
- **Architecture**: EfficientAD (Medium)
- **Accuracy**: 90.0%
- **Inference Time**: ~300-500ms

## Defect Types Detected
- Broken (large)
- Broken (small)
- Contamination

## Usage

```python
from huggingface_hub import hf_hub_download
import torch

# Download model weights
teacher = hf_hub_download(repo_id="MSherbinii/efficientad-bottle", filename="teacher_final.pth")
student = hf_hub_download(repo_id="MSherbinii/efficientad-bottle", filename="student_final.pth")
autoencoder = hf_hub_download(repo_id="MSherbinii/efficientad-bottle", filename="autoencoder_final.pth")
```

## Live Demo
Try it at: https://msherbinii-lumina-ai-detection.hf.space

## Powered by Lumina Tech Platform
Industrial AI quality control system.