lagmoellertim/essentia repository overview

⁠essentia-docker

Docker images for the Essentia audio analysis library

⁠Introduction

Welcome to essentia-docker, a collection of Docker images for Essentia⁠, the open-source C++ library for audio and music analysis, description, and audio-based music information retrieval. This project provides multi-platform Docker images for Essentia, including optional builds with TensorFlow and GPU support. These images are intended as a base for your own development or integration work, rather than as ready-to-use applications.

⁠Features

• Multi-Platform: Images for amd64 and arm64 architectures
• TensorFlow Support: Optional builds with TensorFlow (CPU/GPU)
• Easy Deployment: Run Essentia in a containerized environment with minimal setup
• CI/CD: Automated builds and multi-arch manifests via GitHub Actions

⁠Getting Started

⁠Prerequisites

• Docker⁠ installed on your system

⁠Pulling from Docker Hub

Images are published to Docker Hub under lagmoellertim/essentia.

docker pull docker.io/lagmoellertim/essentia:latest

⁠Available Tags

• latest – Standard Essentia build
• -tensorflow – With TensorFlow (CPU)
• -tensorflow-gpu – With TensorFlow (GPU)

⁠Running Essentia

If you want to use Essentia in your own Docker image, use this as a base:

FROM lagmoellertim/essentia:latest
# ...your build steps...

⁠Building Locally

To build the images yourself:

git clone https://github.com/lagmoellertim/essentia-docker.git
cd essentia-docker
docker build -t essentia:latest .

For TensorFlow builds:

docker build -t essentia:tensorflow --build-arg ENABLE_TENSORFLOW=1 --build-arg TENSORFLOW_USE_GPU=0 .

⁠Contributing

Feel free to open issues or submit pull requests for improvements, new features, or bug fixes!

⁠Author

Tim-Luca Lagmöller (@lagmoellertim⁠)

⁠Donations / Sponsors

I'm part of the official GitHub Sponsors program where you can support me on a monthly basis.

You can also contribute by buying me a coffee (this is a one-time donation).

Thank you for your support!

⁠License

The Code is licensed under the

MIT License⁠

Copyright © 2025-present, Tim-Luca Lagmöller⁠

⁠Have fun :tada: