TremorAI: Advanced Earthquake Detection System

TremorAI is a state-of-the-art deep learning application designed to detect and characterize seismic events in real-time. Developed by Amitoj Singh and Varun Bothra, this system leverages advanced Convolutional Neural Networks (CNNs) to distinguish between earthquake signals and seismic noise with high precision.

🚀 Features

• Real-Time Detection: Instantly analyzes seismic signals to detect potential earthquakes.
• High Accuracy: Powered by a custom-trained CNN model achieving >99% accuracy on the STEAD dataset.
• Magnitude Estimation: Automatically estimates the magnitude of detected events.
• Interactive Dashboard: Modern, dark-themed web interface for easy data visualization and analysis.
• Universal Compatibility: Supports standard .csv and .json seismic data formats.

🛠️ Tech Stack

• Frontend: HTML5, CSS3, JavaScript (Chart.js)
• Backend: Python, FastAPI
• AI Core: TensorFlow, Keras, NumPy
• Data Processing: Pandas, SciPy

🏁 Getting Started

Prerequisites

• Python 3.8 or higher
• Pip package manager

Installation

1. Clone the repository:

   git clone https://github.com/yourusername/TremorAI.git
   cd TremorAI

2. Install dependencies:

   pip install -r requirements.txt

Running the Application

1. Start the Server:

   python backend/main.py

   The server will start at http://0.0.0.0:8000.

2. Access the Dashboard: Open your web browser and navigate to http://localhost:8000.

3. Analyze Data:

   • Upload your seismic data file (.csv or .json).
   • Or click "Use Example Data" to load a sample.
   • Click "Analyze Signal" to see the AI's prediction.

📂 Project Structure

• backend/: API server and model inference logic.
• frontend/: User interface source code.
• core/: Core machine learning and data processing utilities.
• examples/: Sample datasets for testing.
• notebooks/: Research and training experiments.

🧠 Model Performance

TremorAI uses a specialized CNN architecture optimized for seismic signal processing.

Metric	Score
Accuracy	99.01%
Precision	99.27%
Recall	98.83%
F1 Score	99.05%

Note

The model was trained on over 100,000 seismic samples from the STEAD dataset.

👥 Authors

• Amitoj Singh - Lead Developer & AI Researcher
• Varun Bothra - Lead Developer & System Architect

📄 License

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

---

Built with ❤️ by the TremorAI Team