InterGridNet: An Electric Network Frequency Approach for Audio Source Location Classification Using Convolutional Neural Networks

Implementation of InterGridNet, a RawNet-based framework for audio source location classification using Electric Network Frequency (ENF) features, as proposed by Christos Korgialas et al. in InterGridNet: An Electric Network Frequency Approach for Audio Source Location Classification Using Convolutional Neural Networks, presented at SIGNAL 2025.

Table of Contents

• Introduction
• Installation
• Data Preparation
• Model Training
• Model Testing

Introduction

InterGridNet introduces a CNN-based solution for audio source classification using ENF characteristics. The project is designed to handle raw audio recordings, process ENF signals, and classify sources across different grids.

Installation

1. Clone the repository:

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

Data Preparation

1. Place raw audio recordings in the databases/database_raw directory.
2. Split large audio files into smaller segments
3. Normalize and prepare the dataset

Model Training

python train_model_tuner.py

Model Testing

python test_detectFreq.py

Authors

Feel free to send us a message for any issue.

Christos Korgialas (ckorgial@csd.auth.gr)