OpenViSTAI · Agents · Blockchain · Crypto
Submit a project
Menu
Audio & Voice AI

Wave U Net

Wave U Net is a Python implementation of the Wave-U-Net convolutional neural network, which separates audio sources by processing raw waveforms with one-dimensional convolutions and downsampling and upsampling blocks.

Explore the practical guide ↓

Independent OpenViST project profile, based on linked sources.

Wave U Net — OpenViST project profile in Audio & Voice AI.
OpenViST project illustration. See the linked sources for product details.
Understand the project

Practical overview

When it fits

Use it when you need a research-oriented implementation of waveform-based source separation and can work with its documented Python and TensorFlow dependencies.

Who it is for

  • Developers exploring a source-separation implementation for audio tools or workflows.
  • Researchers reproducing the repository’s vocal or multi-instrument separation experiments.

Documented uses

  • Separate vocals from a mixed song using the provided pretrained models.
  • Separate multiple instruments from mixed audio.
  • Train model variants to reproduce the paper’s experiments.
  • Use separated stems in downstream audio-processing workflows.
Before you choose

Requirements & limits

The repository is public and identifies its license as MIT.

  • The implementation uses Python 3.6.8 and TensorFlow 1.8, with libsndfile and CUDA 9 listed as requirements.
  • A GPU is strongly recommended for training, though the README notes that CPU TensorFlow can be used when no GPU is available.
  • Pretrained-model use does not require the training datasets; reproducing experiments requires MUSDB18, and vocal experiments also require CCMixter.
  • The repository points to a separate improved PyTorch implementation; that version is not this repository.

Check current requirements in the official documentation. Source-based guidance does not mean OpenViST has run or independently tested this project.

Source-based practical guide

Install Wave-U-Net Python dependencies

Install the Python packages listed for the Wave-U-Net repository.

Before you start

  • Python 3.6.8
  • libsndfile
  • CUDA 9 for GPU use; the README also allows CPU TensorFlow instead of tensorflow-gpu when no GPU is available.
  • The Wave-U-Net repository cloned locally.

Steps

  1. Install the required system dependencies, including libsndfile and, for GPU use, CUDA 9.

  2. In the downloaded repository directory, install the listed packages with the command below.

    Documented command
    pip install -r requirements.txt

  3. If no GPU is available, use the CPU TensorFlow package instead of tensorflow-gpu.

Expected outcome

The repository's Python package requirements are installed, with TensorFlow selected for GPU or CPU use.

Guide limitations

  • A GPU is strongly recommended to avoid very long training times.
  • The documented environment uses older pinned package versions; no compatibility with other environments is specified.
  • This covers dependency installation only, not running separation or training.
Documentation 1 ↗
Evidence and transparency

Sources & review

Official repository ↗Guide documentation ↗
Source record details
Source availability
Active
Sources checked
5 November 2026
Original website title
f90/Wave-U-Net · Official homepage title · retained in its source language
How we review sourcesSuggest a correction

Related options

Wave-U-Net-Pytorch — A separate improved PyTorch version linked by this repository; it is not the implementation profiled here.