A holistic approach to polyphonic music transcription with neural networks

Miguel A. Román, Antonio Pertusa, Jorge Calvo-Zaragoza

We present a framework based on neural networks to extract music scores directly from polyphonic audio in an end-to-end fashion. Most previous Automatic Music Transcription (AMT) m…