FM_Denoising_DAS

This repository contains code for an ML denoiser that is one of the tasks of an overall foundation model.

Project Overview

The goal of this project is to develop and evaluate methods for denoising DAS signals to improve downstream processing. The approach leverages time-frequency transformations and deep learning to enhance signal quality.

Getting Started

Follow the steps below to set up the project locally.

---

Prerequisites

• Python 3.8 or later
• Git
• conda or virtualenv

---

Clone the Repository

git clone https://github.com/uwfiberlab/FM_Denoising_DAS.git
cd FM_Denoising_DAS

---

Create a new conda environment using the provided env.yml:

conda env create -f env.yml
conda activate fm_denoising_das

Project Structure

FM_Denoising_DAS/
├── data/               # Contains raw or preprocessed DAS data (not included)
├── models/             # Saved model checkpoints and architectures
├── env.yml             # Conda environment definition
└── README.md           # Project documentation

---

Contact

For questions or collaboration inquiries, please reach out to the UW Fiber Lab or open an issue in this repo.