You can not select more than 25 topics Topics must start with a letter or number, can include dashes ('-') and can be up to 35 characters long.
PaddleSpeech/paddleaudio
KP ee9972a0d3
Rename to paddleaudio.
3 years ago
..
examples Rename to paddleaudio. 3 years ago
paddleaudio Rename to paddleaudio. 3 years ago
test Rename to paddleaudio. 3 years ago
.gitignore Rename to paddleaudio. 3 years ago
.pre-commit-config.yaml Rename to paddleaudio. 3 years ago
.style.yapf Rename to paddleaudio. 3 years ago
LICENSE Rename to paddleaudio. 3 years ago
README.md Rename to paddleaudio. 3 years ago
requirements.txt Rename to paddleaudio. 3 years ago
setup.py Rename to paddleaudio. 3 years ago

README.md

PaddleAudio: The audio library for PaddlePaddle

Introduction

PaddleAudio is the audio toolkit to speed up your audio research and development loop in PaddlePaddle. It currently provides a collection of audio datasets, feature-extraction functions, audio transforms,state-of-the-art pre-trained models in sound tagging/classification and anomaly sound detection. More models and features are on the roadmap.

Features

  • Spectrogram and related features are compatible with librosa.
  • State-of-the-art models in sound tagging on Audioset, sound classification on esc50, and more to come.
  • Ready-to-use audio embedding with a line of code, includes sound embedding and more on the roadmap.
  • Data loading supports for common open source audio in multiple languages including English, Mandarin and so on.

Install

git clone https://github.com/PaddlePaddle/models
cd models/PaddleAudio
pip install .

Quick start

Audio loading and feature extraction

import paddleaudio as pa
s,r = pa.load(f)
mel_spect = pa.melspectrogram(s,sr=r)

Examples

We provide a set of examples to help you get started in using PaddleAudio quickly.

Please refer to example directory for more details.