Rapid Scanning of Spectrograms for Efficient Identification of Bioacoustic Events in Big Data

Truskinger, Anthony; Cottman-Fields, Mark; Roe, Paul

doi:10.1109/escience.2013.25

Cited by 21 publications

(13 citation statements)

References 16 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…This is achieved by disabling audio playback, and flashing successive spectrograms past participants at high speeds. We have achieved as much as 12x speedup for time taken to identify a species in a section of audio [7].…”

Section: Semi-automated Approachmentioning

confidence: 92%

See 1 more Smart Citation

Managing and Analysing Big Audio Data for Environmental Monitoring

Zhang

Huang

Cottman-Fields

et al. 2013

2013 IEEE 16th International Conference on Computational Science and Engineering

Self Cite

View full text Add to dashboard Cite

Environmental monitoring is becoming critical as human activity and climate change place greater pressures on biodiversity, leading to an increasing need for data to make informed decisions. Acoustic sensors can help collect data across large areas for extended periods making them attractive in environmental monitoring.However, managing and analysing large volumes of environmental acoustic data is a great challenge and is consequently hindering the effective utilization of the big dataset collected. This paper presents an overview of our current techniques for collecting, storing and analysing large volumes of acoustic data efficiently, accurately, and cost-effectively.

show abstract

Section: Semi-automated Approachmentioning

confidence: 92%

“…The majority of our call recognition algorithms employ features derived from spectrograms [5,7]. The signal is framed using a window of 512 samples (23.2ms) which offers a reasonable compromise between time and frequency resolution.…”

Section: Automated Approachmentioning

confidence: 99%

Managing and Analysing Big Audio Data for Environmental Monitoring

Zhang

Huang

Cottman-Fields

et al. 2013

2013 IEEE 16th International Conference on Computational Science and Engineering

Self Cite

View full text Add to dashboard Cite

show abstract

“…The detection of the species in the audio recordings followed two steps: (1) First author (MCC) listened to all recordings from 05:30 to 10:00 hr, and then, one recording every hour (e.g., 11:00, 12:00, 13:00 hr) until 18:00 hr on the first survey day for each sampling site ( n = 35, 1‐min recordings/site) for a total of 2,100 validated recordings, and (2) with a preliminary bird list from the first survey day from each site, MCC visually scanned all spectrograms of recordings (Campos‐Cerqueira & Aide, ; Truskinger, Cottman‐Fields, Johnson, & Roe, ), from 05:30 to 18:00 hr for all subsequent days, thus an additional approximately 41,000 1 min recordings were evaluated. During this process, MCC listened to every recording that included a previously unidentified vocalization for the site.…”

Section: Methodsmentioning

confidence: 99%

Have bird distributions shifted along an elevational gradient on a tropical mountain?

Campos‐Cerqueira

Arendt

Wunderle

et al. 2017

Ecology and Evolution

View full text Add to dashboard Cite

An upward shift in elevation is one of the most conspicuous species responses to climate change. Nevertheless, downward shifts and, apparently, the absences of response have also been recently reported. Given the growing evidence of multiple responses of species distributions due to climate change and the paucity of studies in the tropics, we evaluated the response of a montane bird community to climate change, without the confounding effects of land‐use change. To test for elevational shifts, we compared the distribution of 21 avian species in 1998 and 2015 using occupancy models. The historical data set was based on point counts, whereas the contemporary data set was based on acoustic monitoring. We detected a similar number of species in historical (36) and contemporary data sets (33). We show an overall pattern of no significant change in range limits for most species, although there was a significant shift in the range limit of eight species (38%). Elevation limits shifted mostly upward, and this pattern was more common for upper than lower limits. Our results highlight the variability of species responses to climate change and illustrate how acoustic monitoring provides an easy and powerful way to monitor animal populations along elevational gradients.

show abstract

“…We deliberately discarded dusk and dawn choruses, at 05:00 and 06:00 h and at 17:00 and 18:00 h, because of the dominance of overlapping bird sounds. We simultaneously listened and visualized the selected recordings using a spectrogram generated with Audacity (non-overlapping 1024 sample Blackman-Harris window), for a more accurate discrimination of call overlaps (Truskinger et al 2013). First, we classified calls-defined here as a distinct acoustic production-in four main acoustic groups as follow: birds, crickets, katydids and others (cicadas, anurans, cats and bats).…”

Section: Calling Identification and Countmentioning

confidence: 99%

Cricket calling communities as an indicator of the invasive ant Wasmannia auropunctata in an insular biodiversity hotspot

et al. 2017

View full text Add to dashboard Cite

Invasive species are a major concern for the maintenance of ecosystem services and biodiversity but are difficult to mitigate. Upstream solutions to prevent their impact, including their detection, are needed. Wasmannia auropunctata, an invasive ant living in vagile supercolonies, is especially hard to track and is a major threat for tropical ecosystems and local animal communities. As part of such tropical communities, crickets are sensitive to ecological conditions, easy to collect, detectable and identifiable through their species-specific calls. Here, we evaluated the use of an acoustic community of crickets as an indicator of the presence of W. auropunctata in New Caledonia. We evaluated the dominance of the crickets in the soundscape, describe the cricket community structure and diversity along a shrubland to forest gradient, characterize these cricket communities structure and diversity in the light of ongoing invasion by W. auropunctata, and identify cricket species' indicators of the invasion. Acoustic recordings collected on 24 sites were described using humanlistening and spectrographic visualization. The results demonstrated a clear dominance of the cricket group in the New Caledonian nocturnal soundscapes. Each habitat harbored a specific acoustic cricket community related to specific environmental attributes including vegetation height, daily variation of humidity and temperature. The presence of W. auropunctata was significantly associated with a lower cricket acoustic activity and species richness at night. Of the 19 species detected, four nocturnal species were identified as indicator of non-invaded forests and preforests. This work supports the use of acoustic as an alternative method to detect invasion.

show abstract

Rapid Scanning of Spectrograms for Efficient Identification of Bioacoustic Events in Big Data

Cited by 21 publications

References 16 publications

Managing and Analysing Big Audio Data for Environmental Monitoring

Managing and Analysing Big Audio Data for Environmental Monitoring

Have bird distributions shifted along an elevational gradient on a tropical mountain?

Cricket calling communities as an indicator of the invasive ant Wasmannia auropunctata in an insular biodiversity hotspot

Contact Info

Product

Resources

About