# Multitask Learning for Polyphonic Piano Transcription, a Case Study

## TL;DR

This paper explores how multitask learning with multiple prediction targets affects the accuracy of polyphonic piano transcription using CNNs on the MAESTRO dataset.

## Contribution

It investigates the impact of adding auxiliary prediction tasks in CNN-based models for polyphonic piano transcription.

## Key findings

- Additional objectives improve transcription accuracy.
- Different CNN architectures respond variably to multitask learning.
- Quantitative analysis on the MAESTRO dataset demonstrates performance differences.

## Abstract

Viewing polyphonic piano transcription as a multitask learning problem, where we need to simultaneously predict onsets, intermediate frames and offsets of notes, we investigate the performance impact of additional prediction targets, using a variety of suitable convolutional neural network architectures. We quantify performance differences of additional objectives on the large MAESTRO dataset.