--- title: Bacteria Detection System emoji: 🦠 colorFrom: blue colorTo: purple sdk: gradio sdk_version: 5.49.1 app_file: app.py pinned: false license: mit short_description: Classifies bacteria images into Gram+/- cocci and bacilli wi --- # 🦠 Bacteria Detection System A deep learning application for detecting and classifying bacteria in microscope images using YOLOv8. ## 🎯 Overview This application automatically detects and classifies bacteria based on: - **Gram staining:** Gram-positive (G+) vs Gram-negative (G-) - **Shape:** Cocci (spherical) vs Bacilli (rod-shaped) ## 🔬 Features - **Real-time Detection:** Upload microscope images and get instant results - **4 Classes:** - G- Cocci (Gram-negative cocci) - G+ Cocci (Gram-positive cocci) - G- Bacilli (Gram-negative bacilli) - G+ Bacilli (Gram-positive bacilli) - **Visual Results:** Color-coded bounding boxes with confidence scores - **Detailed Statistics:** Class distribution and detection counts ## 🚀 How to Use 1. **Upload Image:** Click the upload area or drag & drop a microscope image 2. **Detect:** Click the "🔬 Detect Bacteria" button 3. **View Results:** See detected bacteria with bounding boxes and classification ## 🧬 Model Details - **Architecture:** YOLOv8 Nano - **Training Dataset:** Clinical Bacteria Detection Dataset (6,005 images) - **Performance:** - Precision: ~87% - Recall: ~82% - mAP50: ~87% - mAP50-95: ~64% - **Confidence Threshold:** 25% - **IoU Threshold:** 45% ## 📊 Classes | Class | Description | Gram Stain | Shape | | ---------- | --------------------- | ---------------------- | ---------- | | G- Cocci | Gram-negative cocci | Negative (Pink/Red) | Spherical | | G+ Cocci | Gram-positive cocci | Positive (Purple/Blue) | Spherical | | G- Bacilli | Gram-negative bacilli | Negative (Pink/Red) | Rod-shaped | | G+ Bacilli | Gram-positive bacilli | Positive (Purple/Blue) | Rod-shaped | ## 🛠️ Technology Stack - **Model:** YOLOv8 (Ultralytics) - **Framework:** Gradio - **Backend:** PyTorch - **Image Processing:** OpenCV ## 📖 Clinical Relevance Gram staining and bacterial shape identification are crucial for: - Initial antibiotic selection - Infection diagnosis - Treatment planning - Laboratory screening **Note:** This is a research tool and should not replace professional medical diagnosis. ## 🔗 Links - [Dataset Paper](https://www.nature.com/articles/s41597-024-03370-5) - [YOLOv8 Documentation](https://docs.ultralytics.com/) - [Gradio Documentation](https://gradio.app/) ## 📝 License MIT License ## 🙏 Acknowledgments - Dataset: Clinical Bacteria Detection Dataset (Nature Scientific Data) - Model: YOLOv8 by Ultralytics - Interface: Gradio by Hugging Face --- **Made with ❤️ for medical research and education**