Master 2017 2018
Stages de la spécialité SAR
Automatic drum loop transcription using deep learning


Site : AudioGaming
Lieu : Toulouse
Encadrant : Chunghsin YEH
Dates :du 01/03/2017 au 31/07/2017
Rémunération :554€
Mots-clés : Parcours ATIAM : Acoustique, Parcours ATIAM : Informatique musicale, Parcours ATIAM : Traitement du signal

Description

We will study several deep learning techniques for automatic drum transcription for both acoustic drums and electronic drums. The candidate will carry out the following tasks

  1. Create an automatic database generation/augmentation system
  2. Review state of the art of MIR based on deep learning
  3. Evaluate several neural networks and compare the results against an existing rule-based method.

Bibliographie

Vogl, R. Drum transcription from polyphonic music with recurrent neural networks. ICASSP 2017.