Birdvox-Full-Night: A Dataset and Benchmark for Avian Flight Call Detection

Vincent Lostanlen, Justin Salamon, Andrew Farnsworth, Steve Kelling, Juan Bello

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Abstract

This article addresses the automatic detection of vocal, nocturnally migrating birds from a network of acoustic sensors. Thus far, owing to the lack of annotated continuous recordings, existing methods had been benchmarked in a binary classification setting (presence vs. absence). Instead, with the aim of comparing them in event detection, we release BirdVox-full-night, a dataset of 62 hours of audio comprising 35402 flight calls of nocturnally migrating birds, as recorded from 6 sensors. We find a large performance gap between energy-based detection functions and data-driven machine listening. The best model is a deep convolutional neural network trained with data augmentation. We correlate recall with the density of flight calls over time and frequency and identify the main causes of false alarm.

Original languageEnglish (US)
Title of host publication2018 IEEE International Conference on Acoustics, Speech, and Signal Processing, ICASSP 2018 - Proceedings
PublisherInstitute of Electrical and Electronics Engineers Inc.
Pages266-270
Number of pages5
Volume2018-April
ISBN (Print)9781538646588
DOIs
StatePublished - Sep 10 2018
Event2018 IEEE International Conference on Acoustics, Speech, and Signal Processing, ICASSP 2018 - Calgary, Canada
Duration: Apr 15 2018Apr 20 2018

Other

Other2018 IEEE International Conference on Acoustics, Speech, and Signal Processing, ICASSP 2018
CountryCanada
CityCalgary
Period4/15/184/20/18

Fingerprint

Birds
Sensors
Energy gap
Acoustics
Neural networks

Keywords

  • Acoustic signal detection
  • Audio databases
  • Ecosystems
  • Multi-layer neural network
  • Supervised learning

ASJC Scopus subject areas

  • Software
  • Signal Processing
  • Electrical and Electronic Engineering

Cite this

Lostanlen, V., Salamon, J., Farnsworth, A., Kelling, S., & Bello, J. (2018). Birdvox-Full-Night: A Dataset and Benchmark for Avian Flight Call Detection. In 2018 IEEE International Conference on Acoustics, Speech, and Signal Processing, ICASSP 2018 - Proceedings (Vol. 2018-April, pp. 266-270). [8461410] Institute of Electrical and Electronics Engineers Inc.. https://doi.org/10.1109/ICASSP.2018.8461410

Birdvox-Full-Night : A Dataset and Benchmark for Avian Flight Call Detection. / Lostanlen, Vincent; Salamon, Justin; Farnsworth, Andrew; Kelling, Steve; Bello, Juan.

2018 IEEE International Conference on Acoustics, Speech, and Signal Processing, ICASSP 2018 - Proceedings. Vol. 2018-April Institute of Electrical and Electronics Engineers Inc., 2018. p. 266-270 8461410.

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Lostanlen, V, Salamon, J, Farnsworth, A, Kelling, S & Bello, J 2018, Birdvox-Full-Night: A Dataset and Benchmark for Avian Flight Call Detection. in 2018 IEEE International Conference on Acoustics, Speech, and Signal Processing, ICASSP 2018 - Proceedings. vol. 2018-April, 8461410, Institute of Electrical and Electronics Engineers Inc., pp. 266-270, 2018 IEEE International Conference on Acoustics, Speech, and Signal Processing, ICASSP 2018, Calgary, Canada, 4/15/18. https://doi.org/10.1109/ICASSP.2018.8461410
Lostanlen V, Salamon J, Farnsworth A, Kelling S, Bello J. Birdvox-Full-Night: A Dataset and Benchmark for Avian Flight Call Detection. In 2018 IEEE International Conference on Acoustics, Speech, and Signal Processing, ICASSP 2018 - Proceedings. Vol. 2018-April. Institute of Electrical and Electronics Engineers Inc. 2018. p. 266-270. 8461410 https://doi.org/10.1109/ICASSP.2018.8461410
Lostanlen, Vincent ; Salamon, Justin ; Farnsworth, Andrew ; Kelling, Steve ; Bello, Juan. / Birdvox-Full-Night : A Dataset and Benchmark for Avian Flight Call Detection. 2018 IEEE International Conference on Acoustics, Speech, and Signal Processing, ICASSP 2018 - Proceedings. Vol. 2018-April Institute of Electrical and Electronics Engineers Inc., 2018. pp. 266-270
@inproceedings{6a5d87626688468fae4941168cfdfd30,
title = "Birdvox-Full-Night: A Dataset and Benchmark for Avian Flight Call Detection",
abstract = "This article addresses the automatic detection of vocal, nocturnally migrating birds from a network of acoustic sensors. Thus far, owing to the lack of annotated continuous recordings, existing methods had been benchmarked in a binary classification setting (presence vs. absence). Instead, with the aim of comparing them in event detection, we release BirdVox-full-night, a dataset of 62 hours of audio comprising 35402 flight calls of nocturnally migrating birds, as recorded from 6 sensors. We find a large performance gap between energy-based detection functions and data-driven machine listening. The best model is a deep convolutional neural network trained with data augmentation. We correlate recall with the density of flight calls over time and frequency and identify the main causes of false alarm.",
keywords = "Acoustic signal detection, Audio databases, Ecosystems, Multi-layer neural network, Supervised learning",
author = "Vincent Lostanlen and Justin Salamon and Andrew Farnsworth and Steve Kelling and Juan Bello",
year = "2018",
month = "9",
day = "10",
doi = "10.1109/ICASSP.2018.8461410",
language = "English (US)",
isbn = "9781538646588",
volume = "2018-April",
pages = "266--270",
booktitle = "2018 IEEE International Conference on Acoustics, Speech, and Signal Processing, ICASSP 2018 - Proceedings",
publisher = "Institute of Electrical and Electronics Engineers Inc.",

}

TY - GEN

T1 - Birdvox-Full-Night

T2 - A Dataset and Benchmark for Avian Flight Call Detection

AU - Lostanlen, Vincent

AU - Salamon, Justin

AU - Farnsworth, Andrew

AU - Kelling, Steve

AU - Bello, Juan

PY - 2018/9/10

Y1 - 2018/9/10

N2 - This article addresses the automatic detection of vocal, nocturnally migrating birds from a network of acoustic sensors. Thus far, owing to the lack of annotated continuous recordings, existing methods had been benchmarked in a binary classification setting (presence vs. absence). Instead, with the aim of comparing them in event detection, we release BirdVox-full-night, a dataset of 62 hours of audio comprising 35402 flight calls of nocturnally migrating birds, as recorded from 6 sensors. We find a large performance gap between energy-based detection functions and data-driven machine listening. The best model is a deep convolutional neural network trained with data augmentation. We correlate recall with the density of flight calls over time and frequency and identify the main causes of false alarm.

AB - This article addresses the automatic detection of vocal, nocturnally migrating birds from a network of acoustic sensors. Thus far, owing to the lack of annotated continuous recordings, existing methods had been benchmarked in a binary classification setting (presence vs. absence). Instead, with the aim of comparing them in event detection, we release BirdVox-full-night, a dataset of 62 hours of audio comprising 35402 flight calls of nocturnally migrating birds, as recorded from 6 sensors. We find a large performance gap between energy-based detection functions and data-driven machine listening. The best model is a deep convolutional neural network trained with data augmentation. We correlate recall with the density of flight calls over time and frequency and identify the main causes of false alarm.

KW - Acoustic signal detection

KW - Audio databases

KW - Ecosystems

KW - Multi-layer neural network

KW - Supervised learning

UR - http://www.scopus.com/inward/record.url?scp=85052394269&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85052394269&partnerID=8YFLogxK

U2 - 10.1109/ICASSP.2018.8461410

DO - 10.1109/ICASSP.2018.8461410

M3 - Conference contribution

SN - 9781538646588

VL - 2018-April

SP - 266

EP - 270

BT - 2018 IEEE International Conference on Acoustics, Speech, and Signal Processing, ICASSP 2018 - Proceedings

PB - Institute of Electrical and Electronics Engineers Inc.

ER -