---
license: mit
base_model:
- Ultralytics/YOLO11
datasets:
- JakeTurner616/mtg_face_objects_classifcation
---
<div align="center">
  <h1>MTG Card Segmentation with YOLO11</h1>
</div>

<div align="center">
  <img src="result.jpg" alt="Detection Output" width="400"/>
</div>

**Overview:**  
This repository contains a custom-trained YOLO11 model for segmenting ***high-resolution*** Magic: The Gathering card scans. The model is trained to detect 7 key card elements:  
- **card_title**  
- **card_art**  
- **card_type**  
- **card_set_symbol**  
- **card_mana_cost**  
- **card_oracle_text**  
- **card_power_toughness**

---

## Model Overview

<div style="display: flex; align-items: center;">
  <div style="flex: 1; padding-right: 20px;">
    <p>
      The model accurately locates and segments the critical graphical elements of MTG cards. Trained on annotated high-resolution scans, it can detect these segments in new images with high precision, enabling automated analysis and digital processing of card graphics.
    </p>
  </div>
  <div style="flex: 1; text-align: center;">
    <img src="result.jpg" alt="Detection Output" width="300"/>
  </div>
</div>

---

## Training Results

<div style="display: flex; align-items: center;">
  <div style="flex: 1; text-align: center;">
    <img src="results.png" alt="Training Metrics Graph" width="600"/>
  </div>
</div>

---