Acknowledgements
// Last updated: September 20, 2026
Open Source & Research
Trakx is built on work that other people gave away. This page credits it.
Neural Audio Fingerprinting
Trakx identifies unreleased tracks using a model trained with neural-audio-fp, released under the MIT License.
Copyright © 2021 Sungkyun Chang
The approach is described in:
Sungkyun Chang, Donmoon Lee, Jeongsoo Park, Hyungui Lim, Kyogu Lee, Karam Ko and Yoonchang Han.
Neural Audio Fingerprint for High-specific Audio Retrieval based on Contrastive Learning.
International Conference on Acoustics, Speech and Signal Processing (ICASSP), 2021.
The model running in Trakx was trained by us on our own catalogue. The authors are not affiliated with Trakx and do not endorse it.
Other Components
- ShazamKit — Apple. Used for on-device recognition of released music.
- Firebase — Google. Authentication, database, storage and serverless functions.
- TensorFlow — Apache License 2.0. Runs the recognition model.
- Chromaprint — LGPL 2.1. Bundled from an earlier version of the app and no longer used for recognition.