Human Voice Pitch Estimation: A Convolutional Network with Auto-Labeled and Synthetic Data

TL;DR

This paper introduces a convolutional neural network for human voice pitch estimation that leverages auto-labeled and synthetic data, achieving robust performance across diverse audio datasets in music and voice applications.

Contribution

It presents a novel CNN architecture trained on combined synthetic and auto-labeled data for improved pitch extraction from human singing voices.

Findings

Effective across synthetic and real-world datasets

Outperforms traditional pitch estimation methods

Robust in diverse singing and speech scenarios

Abstract

In the domain of music and sound processing, pitch extraction plays a pivotal role. Our research presents a specialized convolutional neural network designed for pitch extraction, particularly from the human singing voice in acapella performances. Notably, our approach combines synthetic data with auto-labeled acapella sung audio, creating a robust training environment. Evaluation across datasets comprising synthetic sounds, opera recordings, and time-stretched vowels demonstrates its efficacy. This work paves the way for enhanced pitch extraction in both music and voice settings.