praudio provides audio preprocessing framework for Deep Learning audio applications

Related tags

Audiopraudio
Overview

README

praudio provides objects and a script for performing complex preprocessing operations on entire audio datasets with one command.

praudio is implemented having Deep Learning audio/music applications in mind.

Operations are carried out on CPU. Preprocessing can also be run on-the-fly, for example, while training a model.

The library uses librosa as an audio processing backend.

How do I install the library?

You can install praudio both with pip via PyPi, and by cloning the praudio repo from GitHub.

For both approaches, it's advisable to use a dedicated Python virtual environment.

Installing from PyPi

Installing from PyPi is the easiest option. In the terminal type:

$ pip install praudio

Installing from GitHub

First, you should clone the repository from GitHub:

$ git clone [email protected]:musikalkemist/praudio.git

Then, move to the project root and, to install the package, type in the terminal:

$ pip install .

You can also use a rule in the available Makefile (see below):

$ make install 

To install the package in development mode use:

$ pip install -e .[testing]

You can also use a rule in Makefile:

$ make install_dev 

This will install all the packages necessary to run the tests, lint, type checker. It will also install the package in 'editable' mode, which is ideal for development.

Python version

praudio works in Python 3.6, 3.7, 3.8.

How do I preprocess an audio dataset?

The core of the library is the preprocess entry point. This script works with a config file. You set the type of preprocessing you want to apply in a yaml file, and then run the script. Your dataset will be entirely preprocessed and the results recursively stored in a directory of your choice that can potentially be created from scratch.

To run the entry point, ensure the library is installed and then type:

$ preprocess /path/to/config.yml

In the config.yml, you should provide the following parameters:

  • dataset_dir: Path to the directory where your audio dataset is stored
  • save_dir: Path where to save the preprocessed audio.
  • Under file_preprocessor, you should provide settings for loader and transforms_chain.
  • loader: Provide settings for the loader.
  • transforms_chain: Parameters for each transform in the sequence. of transforms which are applied to your data (i.e., TransformChain).

These config parameters are used to dinamically initialise the relative objects in the library. To learn what parameters are available at each level in the config file, please refer to the docstrings in the relative objects.

Check out test/config.sampleconfig.yml to see an example of a valid config file.

Package structure

The package is divided into a number of subpackages:

  • config
  • creation
  • io
  • preprocessors
  • transforms

config has facilities to load, save, and validate configuration files, which are used to specify the types of preprocessing pipelines to use.

creation has classes that are responsible to instantiate key objects in the library.

io contains facilities to load / save audio signals from / to files.

preprocessors features objects which are responsible to preprocess single audio files, from loading to storing, as well as, batch of files.

transforms contains a series of objects which manipulate audio signals, such as short-time Fourier transform, log, scaling.

What's the Makefile for?

The Makefile has a series of rules that can be used to ensure quality of the code, and automate repetitive tasks.

Linter

The project uses pylint. The linter helps enforcing a coding standard, sniffs for code smells and offers simple refactoring suggestions.

To run the linter type:

$ make lint

Typehint

The project uses mypy. mypy is an optional static type checker for Python. You can add type hints (PEP 484) to your Python programs, and use mypy to type check them statically.

To run the type checker type:

$ make typehint

Testing

The project uses pytest for unittests. Tests can be run in one go using coverage. This package suggests the percentage of code that is covered in unittests.

To run all the unittests type:

$ make test

Checklist

Checklist is a utility rule that runs the linter, type checker, and the test suite in one go:

$ make checklist

Clean

Use the clean rule to get rid of pyc files and __pychache__:

$ make clean

Dependencies

praudio has the following dependencies:

  • librosa==0.8.1
  • pyyaml==5.4.1
  • types-PyYAML==5.4.6

librosa is extensively used to extract audio features in transform objects.

Current limitations

The praudio preprocessors are capable of operating only on mono signals. This is a significant limitation if you are working in generative music. If you are using the library for audio / music analysis, this shouldn't be a problem.

Future improvements

  • Add audio augmentation / padding / cropping transforms.
  • Enable preprocessing of signals with multiple channels.
  • Turn transform parameters into full-fledged objects (e.g., STFTParams)
  • Instead of using a dictionary for configurations, instantiate parameter objects with validation
  • Implement different types of Savers / Loaders with factories to produce them.
Owner
Valerio Velardo
AI audio/music researcher. Love Python.
Valerio Velardo
Real-time audio visualizations (spectrum, spectrogram, etc.)

Friture Friture is an application to visualize and analyze live audio data in real-time. Friture displays audio data in several widgets, such as a sco

Timothée Lecomte 700 Dec 31, 2022
🎵 A repository for manually annotating files to create labeled acoustic datasets for machine learning.

🎵 A repository for manually annotating files to create labeled acoustic datasets for machine learning.

Jim Schwoebel 28 Dec 22, 2022
Audio augmentations library for PyTorch for audio in the time-domain

Audio augmentations library for PyTorch for audio in the time-domain, with support for stochastic data augmentations as used often in self-supervised / contrastive learning.

Janne 166 Jan 08, 2023
Linear Prediction Coefficients estimation from mel-spectrogram implemented in Python based on Levinson-Durbin algorithm.

LPC_for_TTS Linear Prediction Coefficients estimation from mel-spectrogram implemented in Python based on Levinson-Durbin algorithm. 基于Levinson-Durbin

Zewang ZHANG 58 Nov 17, 2022
Improved Python UI to convert Youtube URL to .mp3 file.

YT-MP3 Improved Python UI to convert Youtube URL to .mp3 file. How to use? Just run python3 main.py Enter the URL of the video Enter the PATH of where

8 Jun 19, 2022
Simple discord bot by @merive 🤖

Parzibot Powerful and Useful Discord Bot on Python. The source code of the bot is available to everyone. Parzibot uses English language. This is free

merive_ 3 Dec 28, 2022
Terminal-based audio-to-text converter

att Terminal-based audio-to-text converter Project description A terminal-based audio-to-text converter written in python, enabling you to convert .wa

Sven Eschlbeck 4 Dec 15, 2022
Audio fingerprinting and recognition in Python

dejavu Audio fingerprinting and recognition algorithm implemented in Python, see the explanation here: How it works Dejavu can memorize audio by liste

Will Drevo 6k Jan 06, 2023
Speech recognition module for Python, supporting several engines and APIs, online and offline.

SpeechRecognition Library for performing speech recognition, with support for several engines and APIs, online and offline. Speech recognition engine/

Anthony Zhang 6.7k Jan 08, 2023
python wrapper for rubberband

pyrubberband A python wrapper for rubberband. For now, this just provides lightweight wrappers for pitch-shifting and time-stretching. All processing

Brian McFee 106 Nov 28, 2022
A Python library and tools AUCTUS A6 based radios.

A Python library and tools AUCTUS A6 based radios.

Jonathan Hart 6 Nov 23, 2022
Python CD-DA ripper preferring accuracy over speed

Whipper Whipper is a Python 3 (3.6+) CD-DA ripper based on the morituri project (CDDA ripper for *nix systems aiming for accuracy over speed). It star

671 Jan 04, 2023
Deep learning transformer model that generates unique music sequences.

music-ai Deep learning transformer model that generates unique music sequences. Abstract In 2017, a new state-of-the-art was published for natural lan

xacer 6 Nov 19, 2022
Mopidy is an extensible music server written in Python

Mopidy Mopidy is an extensible music server written in Python. Mopidy plays music from local disk, Spotify, SoundCloud, Google Play Music, and more. Y

Mopidy 7.6k Jan 05, 2023
Neural building blocks for speaker diarization: speech activity detection, speaker change detection, overlapped speech detection, speaker embedding

⚠️ Checkout develop branch to see what is coming in pyannote.audio 2.0: a much smaller and cleaner codebase Python-first API (the good old pyannote-au

pyannote 2.1k Dec 31, 2022
Noinoi music is smoothly playing music on voice chat of telegram.

NOINOI MUSIC BOT ✨ Features Music & Video stream support MultiChat support Playlist & Queue support Skip, Pause, Resume, Stop feature Music & Video do

2 Feb 13, 2022
Port Hitsuboku Kumi Chinese CVVC voicebank to deepvocal. / 筆墨クミDeepvocal中文音源

Hitsuboku Kumi (筆墨クミ) is a UTAU virtual singer developed by Cubialpha. This project ports Hitsuboku Kumi Chinese CVVC voicebank to deepvocal. This is the first open-source deepvocal voicebank on Gith

8 Apr 26, 2022
Gammatone-based spectrograms, using gammatone filterbanks or Fourier transform weightings.

Gammatone Filterbank Toolkit Utilities for analysing sound using perceptual models of human hearing. Jason Heeris, 2013 Summary This is a port of Malc

Jason Heeris 188 Dec 14, 2022
Code to work with wave files!

Code to work with wave files!

Mohammad Dori 3 Jul 15, 2022
Guide & Examples to create deeplearning gstreamer plugins and use them in your pipeline

upai-gst-dl-plugins Guide & Examples to create deeplearning gstreamer plugins and use them in your pipeline Introduction Thanks to the work done by @j

UPAI.IO 11 Dec 11, 2022