SDAIASJun 1, 2021

Omnizart: A General Toolbox for Automatic Music Transcription

arXiv:2106.00497v136 citations
Originality Synthesis-oriented
AI Analysis

This addresses the problem for researchers and practitioners in music information retrieval by offering a general toolbox, though it is incremental as it builds on existing deep learning methods.

The authors tackled the lack of a comprehensive and user-friendly tool for automatic music transcription by releasing Omnizart, a Python library that provides a streamlined solution covering a wide range of instruments and related tasks, resulting in the first such toolkit with models for solo, ensemble, percussion, vocal, chord recognition, and beat/downbeat tracking.

We present and release Omnizart, a new Python library that provides a streamlined solution to automatic music transcription (AMT). Omnizart encompasses modules that construct the life-cycle of deep learning-based AMT, and is designed for ease of use with a compact command-line interface. To the best of our knowledge, Omnizart is the first transcription toolkit which offers models covering a wide class of instruments ranging from solo, instrument ensembles, percussion instruments to vocal, as well as models for chord recognition and beat/downbeat tracking, two music information retrieval (MIR) tasks highly related to AMT.

Code Implementations1 repo
Foundations

The foundational work for this paper's niche, ranked by how specifically the neighbourhood builds on it — not by global fame.

Your Notes