Pitch tracking of bird vocalizations and an automated process using YIN-bird

Colm O’Reilly,Naomi Harte,Hynek Burda

doi:10.1080/23312025.2017.1322025

Abstract

Pitch or fundamental frequency is an important feature of bird song, from which scientists can learn much about a population. To use pitch as a feature, researchers need confidence in their pitch extraction system. Pitch detection algorithms (PDAs) proven to work on human speech may not be suitable for all types of bird vocalizations. This paper discusses pitch estimation performance on a variety of common bird vocalizations. The presence of multiple partials or tones simultaneously, extended frequency sweeps through multiple octaves, and rapid pitch modulations are just some of the difficulties encountered when estimating the pitch of bird song. Carefully tuned parameters improve pitch tracking with YIN, but optimal parameters can change quickly even within one song. YIN is a PDA which estimates pitch of human speech very well. This paper presents YIN-bird, a modified version of YIN which exploits spectrogram properties to automatically set a minimum fundamental frequency parameter for YIN. Gross pitch e...

Highlights

The ability to automatically analyze bird vocalizations would greatly benefit zoologists in their behavioral and ecological studies
Work in this paper reports improvements in pitch tracking of bird vocalizations which is important for quantifying difference in calls or songs of different bird populations
Bird vocalizations may sound no more complex than human speech, but recordings are usually subject to adverse conditions such as contaminant vocalizations and non-homogeneous noise backgrounds (Kogan & Margoliash, 1998)

Summary

Introduction

The ability to automatically analyze bird vocalizations would greatly benefit zoologists in their behavioral and ecological studies. In the last few years, the speech processing community has researched many issues in bird vocalizations, notably species classification (Connor et al, 2012; Fagerlund & Laine, 2014; Graciarena, Delplanche, Shriberg, Stolcke, & Ferrer, 2010; Heller & Pinezich, 2008; Trifa et al, 2008), syllable or phrase classification (Anderson, Dave, & Margoliash, 1996; Chen & Maher, 2006; Tan, Alwan, Kossan, Cody, & Taylor, 2015; Kaewtip, Tan, Alwan & Taylor, 2013; Kogan & Margoliash, 1998; Ranjard & Ross, 2008; Tan, Kaewtip, Cody, Taylor, & Alwan, 2012), and song structure analysis (Lachlan et al, 2013; Sasahara, Cody, Cohen, & Taylor, 2012). The use of songs and calls to delimit species and monitor populations has several practical advantages, e.g. ease and economy of sound recording and analysis (Remsen, 2005)

Objectives

Methods

Conclusion