Welcome to Soundmentations Documentation!

Soundmentations is a Python library for audio data augmentation and sound classification. It provides a collection of audio transforms that can be used to augment audio datasets for machine learning applications.

Contents:

Features

  • Audio Transforms: Pitch shifting, fading, limiting, and trimming

  • Composition: Chain multiple transforms together

  • Probability Control: Apply transforms with specified probabilities

  • Easy Integration: Works seamlessly with NumPy arrays

  • Extensible: Simple base classes for creating custom transforms

Quick Example

import numpy as np
from soundmentations import PitchShift, FadeIn, Compose

# Create a composition of transforms
transform = Compose([
    PitchShift(semitones=2.0, p=0.8),
    FadeIn(duration=0.1, p=0.5),
])

# Apply to audio
audio = np.random.randn(44100)  # 1 second of audio
augmented = transform(audio, sample_rate=44100)

Indices and Tables