bioaug 0.1.1

A toolbox for biosignal augmentation written in Python

BioAug - Biosignal Augmentation in Python

A toolbox for biosignal augmentation written in Python.

Deep learning models achieve remarkable performance with the aid of massive data. This tool allows you to customly augment your biosignals.

Table of Contents

• Installation
• Introduction
• Uasge
• Result

Installation

We strongly recommend the usage of Anaconda for managing your python environments. Clone repo and install requirements.txt in a Python>=3.8.0 environment, including This set-up was tested under Windows 10 and Ubuntu 20.04.

  $ pip install bioaug

Introduction

• bio/GaussianNoise.py
  • class GaussianNoise: Add specified Gaussian noise to biosignals.

• bio/LocalJittering.py
  • class LocalJittering: Add noise of specific length and frequency at random locations in the biosignals.

• bio/RandomCutout.py
  • class RandomCutout: crop a specific length of a biosignals at a random location to simulate signal loss.

• bio/ImpedanceVariation.py
  • class ImpedanceVariation: simulate changes in biosignals during variation in skin impedance.

• bio/Distortion.py
  • class Distortion: simulate biosignals with distortion during real-world use.

• bio/Drift.py
  • class SignalDrift: simulate biosignals with drift during real-world use.

• bio/MagnitudeWarping.py
  • class MagnitudeWarping: apply magnitude warping to biosignals.

• bio/TimeWarping.py
  • class TimeWarping: apply time warping to biosignals.

• bio/Permutation.py
  • class Permutation: apply permutation to biosignals.

• bio/Scaling.py
  • class Scaling: apply scaling to biosignals.

Usage

This library provides a series of data augmentation methods. The following is an example of using the LocalJittering method. The LocalJittering method is used to randomly add local high-frequency jitter to the input time series signal.

Parameters

• p (float) : Probability of applying jitter, range [0, 1].
• alpha (float, tuple, list) : Scale factor for the high-frequency noise. Can be a fixed value, a range, or a list.
• frequency (int, tuple, list) : Frequency of jitter noise. Can be a fixed value, a range, or a list.
• duration (int, tuple, list) : Duration (in time steps) of each jitter event. Can be a fixed value, a range, or a list.
• num_jitters (int, tuple, list) : Number of jitter events to add. Can be a fixed value, a range, or a list.

Example Usage

1. Fixed values, tuples, and lists combination:

from bioaug import LocalJittering

# Example of mixing fixed values, ranges (tuples), and lists
jitter = LocalJittering(
    p=0.8,                  # Probability of applying jitter
    alpha=0.5,              # Fixed alpha value
    frequency=(40, 60),     # Random frequency in the range [40, 60]
    duration=[10, 20, 30],  # Random duration chosen from the list [10, 20, 30]
    num_jitters=2           # Fixed number of jitter events
)

# Assume 'signal' is the input time-series data
augmented_signal = jitter(signal)