BLIP CAM:Self Hosted Live Image Captioning with Real-Time Video Stream 🎥

This repository implements real-time image captioning using the BLIP (Bootstrapped Language-Image Pretraining) model. The system captures live video from your webcam, generates descriptive captions for each frame, and displays them in real-time along with performance metrics.

🚀 Features

• Real-Time Video Processing: Seamless webcam feed capture and display with overlaid captions
• State-of-the-Art Captioning: Powered by Salesforce's BLIP image captioning model (blip-image-captioning-large)
• Hardware Acceleration: CUDA support for GPU-accelerated inference
• Performance Monitoring: Live display of:
  • Frame processing speed (FPS)
  • GPU memory usage
  • Processing latency
• Optimized Architecture: Multi-threaded design for smooth video streaming and caption generation

📋 Requirements

• Python 3.8+
• NVIDIA GPU (optional, for CUDA acceleration)
• Webcam

Core Dependencies

opencv-python>=4.5.0
torch>=1.9.0
transformers>=4.21.0
Pillow>=8.0.0

🛠️ Installation

1. Clone the repository:

git clone https://github.com/zawawiAI/BLIP_CAM.git
cd BLIP_CAM

2. Install dependencies:

pip install -r requirements.txt

3. Run the application:

python BLIP_CAM.py

💡 Use Cases

• Accessibility Tools: Real-time scene description for visually impaired users
• Content Analysis: Automated video content understanding and tagging
• Smart Conferencing: Enhanced video calls with automatic scene descriptions
• Educational Tools: Visual learning assistance and scene comprehension
• Security Systems: Intelligent surveillance with scene description capabilities

🎮 Usage Controls

• Press Q to quit the application
• Press S to save the current frame with caption
• Press P to pause/resume caption generation

🔧 Configuration

The application can be customized through the following parameters in config.py:

• Frame processing resolution
• Caption update frequency
• GPU memory allocation
• Model confidence threshold
• Display preferences

📝 License

This project is licensed under the MIT License - see the LICENSE file for details.

🙏 Acknowledgments

• Salesforce for the BLIP model
• PyTorch team for the deep learning framework
• Hugging Face for the transformers library

📧 Contact

For questions and support, please open an issue in the GitHub repository or reach out to the maintainers.

---

⭐ If you find this project useful, please consider giving it a star!

blip-2025-01-02_09.online-video-cutter.com.2.1.mp4