BSS Maitri - Bharatiya Space Station AI Assistant 🚀

🌟 Overview

BSS Maitri is a comprehensive multimodal AI assistant designed for the Bharatiya Space Station (BAS) to monitor and support crew emotional and physical well-being through advanced audio-video analysis. Built on Google's Gemma 3 model architecture, it provides real-time psychological support and intervention capabilities.

🎯 Problem Statement

Background: Crew members on-board space stations face isolation, sleep disruption, tight schedules and physical discomforts which can trigger psychological & physical issues. Early intervention can prevent errors and serious health issues.

Objective: Develop a multimodal AI assistant for detecting emotional and physical well-being of crew using audio-video inputs.

✨ Key Features

• 🧠 Multimodal Emotion Detection: Real-time analysis of audio (voice tone) and visual (facial expressions) cues
• 🤖 AI-Powered Conversations: Gemma 3-based empathetic responses with cultural sensitivity
• ⚡ Real-time Monitoring: Continuous assessment of crew psychological state
• 🚨 Intelligent Alerts: Automatic detection of critical emotional states with intervention recommendations
• 📱 Offline Operation: Standalone system requiring no internet connectivity
• 🇮🇳 Cultural Adaptation: Tailored for Indian cultural values and communication styles

🎨 Emotion Detection Capabilities

Audio Analysis

• Voice tone and pitch analysis
• Speech pattern recognition
• Stress indicators in vocal characteristics
• Fatigue detection through speech patterns

Visual Analysis

• Facial expression recognition using MediaPipe
• Eye movement and blink pattern analysis
• Facial landmark detection for emotion mapping
• Real-time facial symmetry analysis

Supported Emotional States

• 😌 Calm/Neutral
• 😰 Stress
• 😟 Anxiety
• 😴 Fatigue
• 😢 Sadness
• 😠 Anger
• 😨 Fear
• 😊 Joy
• 😲 Surprise

🚀 Quick Start

Installation

# Clone the repository
git clone https://github.com/TheAnanta/bss-maitri.git
cd bss-maitri

# Install dependencies
pip install -r requirements.txt

# Install the package
pip install -e .

Basic Usage

from bss_maitri import MaitriAssistant

# Initialize the assistant
with MaitriAssistant() as assistant:
    # Text conversation
    response = assistant.process_text_input("Hello Maitri, how are you?")
    print(response['response'])
    
    # Audio emotion detection
    audio_result = assistant.process_audio_input(audio_data)
    print(f"Detected: {audio_result['emotion_analysis']['dominant_emotion']}")

Command Line Interface

# Interactive text mode
python -m bss_maitri.main --mode interactive

# Audio analysis mode
python -m bss_maitri.main --mode audio --audio-file audio.wav

# Video analysis mode (webcam)
python -m bss_maitri.main --mode video --video-source 0

🏗️ Architecture

BSS Maitri AI Assistant
├── 🧠 Core Components
│   ├── MaitriAssistant (Main orchestrator)
│   ├── EmotionDetector (Multimodal fusion)
│   └── ConversationEngine (AI responses)
├── 🎵 Audio Processing
│   └── AudioEmotionDetector (Voice analysis)
├── 👁️ Vision Processing  
│   └── VisionEmotionDetector (Facial analysis)
├── 🤖 AI Models
│   └── GemmaModel (Language model integration)
└── ⚙️ Configuration & Utils
    ├── Config management
    └── Utility functions

🔧 Configuration

Customize the system through bss_maitri/config/config.yaml:

model:
  name: "google/gemma-2-2b-it"
  device: "cuda"  # or "cpu"
  precision: "fp16"

emotion_detection:
  audio:
    sample_rate: 16000
    frame_length: 2048
  vision:
    face_detection_confidence: 0.5
    emotion_threshold: 0.6

conversation:
  max_history: 10
  intervention_threshold: 0.8
  support_mode: "adaptive"

📊 Monitoring & Alerts

The system automatically monitors for:

• 🚨 Critical Emotional States: High stress, severe anxiety
• 😴 Fatigue Levels: Dangerous fatigue affecting mission safety
• ⏱️ Sustained Negative Emotions: Prolonged stress/anxiety periods
• 📞 Communication Gaps: Extended periods without interaction

🛡️ Space Environment Considerations

• Low-latency Processing: Real-time emotion detection and response
• Resource Optimization: Efficient operation on limited computational resources
• Isolation Support: Addresses psychological challenges of space isolation
• Mission Integration: Balances support with mission requirements
• Cultural Sensitivity: Incorporates Indian values and communication styles

📁 Project Structure

bss-maitri/
├── bss_maitri/           # Main package
│   ├── audio/            # Audio emotion detection
│   ├── vision/           # Visual emotion detection  
│   ├── models/           # AI model integration
│   ├── core/             # Core components
│   ├── config/           # Configuration management
│   ├── utils/            # Utility functions
│   └── main.py           # Entry point
├── examples/             # Usage examples
├── tests/                # Test suite
├── docs/                 # Documentation
└── requirements.txt      # Dependencies

🧪 Testing

# Run basic tests
python -m pytest tests/

# Run specific test
python tests/test_basic.py

# Example usage
python examples/basic_usage.py

📚 Documentation

• Complete Documentation - Detailed API reference and guides
• Configuration Guide - System configuration options
• Examples - Code examples and demos

🤝 Contributing

We welcome contributions! Please see our contributing guidelines and submit pull requests for any improvements.

📄 License

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

🙏 Acknowledgments

• Google for the Gemma language model
• ISRO for space mission insights and requirements
• MediaPipe for computer vision capabilities
• Open Source Community for tools and frameworks

📞 Support

• 🐛 Issues: Report bugs via GitHub Issues
• 💬 Discussions: Join our community discussions
• 📧 Contact: Reach out to the development team

---

⚠️ Disclaimer: This system provides psychological support and monitoring capabilities. It should be used as part of a comprehensive crew health monitoring system and does not replace professional medical care.

🚀 Made with ❤️ for the Bharatiya Space Station Mission