Comparative Analysis of Adaptive Filtering and Neural Network-Based Approaches for Audio Signal Denoising

Signal, Image and Video - University of Trento

M. Prosperi

This project investigates the denoising of audio signals through the application of adaptive filtering techniques, including Least Mean Squares (LMS) and Normalized Least Mean Squares (NLMS), alongside a neural network-based approach. The study is divided into two main phases to systematically evaluate the performance of these methods.

The first phase involves the use of a synthetic audio signal, where each denoising method is assessed both qualitatively (through visual inspection of the results) and quantitatively using objective metrics such as Mean Squared Error (MSE), Root Mean Squared Error (RMSE), and the coefficient of determination (R²). This controlled environment allows for a rigorous comparison of the models’ effectiveness under well-defined conditions.

The second phase extends the analysis to real-world audio signals, utilizing the NOIZEUS Speech Corpus (available at https://ecs.utdallas.edu/loizou/speech/noizeus/), a dataset that provides speech signals with various noise conditions. By applying the denoising techniques to this dataset, the project evaluates the practical applicability and robustness of each method in real-world scenarios.

How to Run the Project

Option 1: Run with Google Colab (Recommended)

The fastest way to run the project is by using Google Colab.
Simply upload the provided notebook to Colab, and you can execute it without needing to install any dependencies locally.

Dataset

Before running the project, you'll need to download the dataset and upload it to Colab.

You can download the clean and noisy audio datasets from the following links:

• Clean Audios
• Noisy Audios

Once downloaded, upload the datasets to your Colab environment to begin the processing.

Option 2: Run Locally

If you prefer to run the project on your local machine, ensure you have the following dependencies installed:

Step 1: Clone the Repository

Clone the repository from GitHub to your local machine:

git clone https://github.com/marcopros/siv_project.git
cd siv_project

Step 2: Install dependencies

• Python 3.8+ (recommended)
• Python Libraries:
  • os (built-in module, no need to install)
  • torch (PyTorch)
  • torch.nn (from PyTorch)
  • torch.optim (from PyTorch)
  • numpy
  • matplotlib
  • scikit-learn
  • scipy

Installation Instructions

To install the required libraries, you can use the following command:

pip install torch numpy matplotlib scikit-learn scipy

Step 3: Run the code

Start Jupyter Notebook:

jupyter notebook

Open main.ipynb and run the cells sequentially.