beardless47/swu-card-scanner

SWU Card Scanner - Hugging Face Inference Endpoint

AI-powered card recognition for Star Wars Unlimited TCG.

Model

Uses openai/clip-vit-large-patch14 for image embeddings with a pre-computed database of 262 card embeddings.

Usage

Via Python

import requests
import base64

# Read image
with open("card.jpg", "rb") as f:
    image_b64 = base64.b64encode(f.read()).decode()

# Call endpoint
response = requests.post(
    "https://api-inference.huggingface.co/models/YOUR_USERNAME/swu-card-scanner",
    headers={"Authorization": "Bearer YOUR_HF_TOKEN"},
    json={
        "inputs": image_b64,
        "parameters": {"top_k": 5}
    }
)

print(response.json())

Via cURL

curl -X POST \
  https://api-inference.huggingface.co/models/YOUR_USERNAME/swu-card-scanner \
  -H "Authorization: Bearer YOUR_HF_TOKEN" \
  -H "Content-Type: application/json" \
  -d '{"inputs": "BASE64_IMAGE_DATA"}'

Response Format

{
  "success": true,
  "quality": "excellent",
  "message": "Match confidence: 98.5%",
  "best_match": {
    "card_id": "card_11",
    "confidence": 0.985,
    "name": "Allegiant General Pryde",
    "image_url": "https://cdn.swu-db.com/images/cards/JTL/133.png",
    "data": { ... }
  },
  "alternatives": [ ... ]
}

Files

• handler.py - Custom inference handler
• embeddings_clip-large.pkl - Pre-computed card embeddings
• card_data.json - Card metadata