README.md

metadata

title: WBC Classification
emoji: 🩸
colorFrom: red
colorTo: blue
sdk: docker
pinned: false
license: mit
app_port: 7860

WBC Classification System

An AI-powered White Blood Cell (WBC) classification system that can identify different types of white blood cells from microscopic images.

Features

• Real-time Classification: Upload images and get instant predictions
• Batch Processing: Process multiple images simultaneously
• High Accuracy: Deep learning model trained on medical imagery
• User-friendly Interface: Simple web interface for easy interaction

Supported Cell Types

• Basophil
• Eosinophil
• Lymphocyte
• Monocyte
• Neutrophil

Usage

1. Single Image: Upload a single microscopic image for classification
2. Batch Processing: Upload multiple images for simultaneous processing
3. View Results: Get detailed predictions with confidence scores

API Endpoints

• GET / - Main web interface
• POST /api/predict - Single image prediction
• POST /api/predict_batch - Batch image prediction
• GET /api/health - Health check
• GET /api/model_info - Model information

Technical Details

• Framework: FastAPI with TensorFlow
• Model: Custom CNN architecture
• Input: 128x128 RGB images
• Output: Classification with confidence scores

Development

Built for medical image analysis and research purposes. The model is trained on high-quality microscopic images of white blood cells.

License

MIT License - See LICENSE file for details.