Open Access Repository

Scalable preprocessing of high volume environmental acoustic data for bioacoustic monitoring


Downloads per month over past year

Brown, AS, Garg, SK ORCID: 0000-0003-3510-2464 and Montgomery, J ORCID: 0000-0002-5360-7514 2018 , 'Scalable preprocessing of high volume environmental acoustic data for bioacoustic monitoring' , PLoS One, vol. 13, no. 8 , pp. 1-24 , doi: 10.1371/journal.pone.0201542.

127424.pdf | Download (3MB)

| Preview


In this work, we examine the problem of efficiently preprocessing and denoising highvolume environmental acoustic data, which is a necessary step in many bird monitoringtasks. Preprocessing is typically made up of multiple steps which are consideredseparately from each other. These are often resource intensive, particularly because thevolume of data involved is high. We focus on addressing two challenges within thisproblem: how to combine existing preprocessing tasks while maximising theeffectiveness of each step, and how to process this pipeline quickly and efficiently, sothat it can be used to process high volumes of acoustic data. We describe a distributedsystem designed specifically for this problem, utilising a master-slave model with dataparallelisation. By investigating the impact of individual preprocessing tasks on eachother, and their execution times, we determine an efficient and accurate order forpreprocessing tasks within the distributed system. We find that, using a single core, ourpipeline executes 1.40 times faster compared to manually executing all preprocessingtasks. We then apply our pipeline in the distributed system and evaluate itsperformance. We find that our system is capable of preprocessing bird acousticrecordings at a rate of 174.8 seconds of audio per second of real time with 32 cores over8 virtual machines, which is 21.76 times faster than a serial process.

Item Type: Article
Authors/Creators:Brown, AS and Garg, SK and Montgomery, J
Keywords: cloud Computing, distributed system, big data, bioacoustics
Journal or Publication Title: PLoS One
Publisher: Public Library of Science
ISSN: 1932-6203
DOI / ID Number: 10.1371/journal.pone.0201542
Copyright Information:

Copyright 2018 Brown et al. Licensed under Creative Commons Attribution 4.0 International (CC BY 4.0)

Related URLs:
Item Statistics: View statistics for this item

Actions (login required)

Item Control Page Item Control Page