Automated annotation of Orthoptera songs: first results from analysing the DORSA sound repository

Riede, Klaus; Nischk, Frank; Dietrich, Christian; Thiel, Christian; Schwenker, Friedhelm; Forschungsmuseum, Zoologisches; Koenig, Alexander

doi:10.1665/1082-6467(2006)15[105:aaoosf]2.0.co;2

Cited by 14 publications

(12 citation statements)

References 7 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Automatic acoustic identification systems for Orthoptera are not new (Chesmore & Ohya ; Riede et al . ; Ganchev, Potamitis & Fakotakis ; Lehmann et al . ), but despite these, regional and national monitoring programmes for Orthoptera still largely rely on unstructured opportunistic sampling.…”

Section: Introductionmentioning

confidence: 99%

Potential for coupling the monitoring of bush‐crickets with established large‐scale acoustic monitoring of bats

et al. 2017

View full text Add to dashboard Cite

Summary Monitoring biodiversity over large spatial and temporal scales is crucial for assessing the impact of global changes and environmental mitigation measures. However, large‐scale monitoring of invertebrates remains poorly developed despite the importance of these organisms in ecosystem functioning. Exciting possibilities applicable to professional and citizen science are offered by new recording techniques and methods of semi‐automated species recognition based on sound detection. Static broad‐spectrum detectors deployed to record throughout whole nights have been recommended for standardised acoustic monitoring of bats, but they have the potential to also collect acoustic data for other species groups. Large‐scale deployment of such systems is only viable when combined with robust automated species identification algorithms. Here we examine the potential of such a system for detecting, identifying and monitoring bush‐crickets (Orthoptera of the family Tettigoniidae). We use incidental sound recordings generated by an extensive citizen science bat survey and recordings from intensive site surveys to test a semi‐automated step‐wise method with a classifier for assigning species identities. We assess species’ diel activity patterns to make recommendations for survey timing and interpretation of existing nocturnal data sets and consider the feasibility of determining site occupancy. Of six species of bush‐crickets, the species classifier achieved over 85% accuracy for three, speckled bush‐cricket, dark bush‐cricket and Roesel's bush‐cricket. It should be possible to automatically scan recordings for these species with minimal manual validation. Further refinement of the classifier is required for the three remaining species, in particular for the acoustically similar short‐winged conehead and long‐winged conehead. Diel activity patterns are species specific and it may be necessary to adjust the hours over which the detectors record to increase detection of key species, but this must be weighed against the costs in terms of increased memory and battery use and equipment security during daytime. We conclude that with logistical support and centralised semi‐automated species identification it is now possible for the public to contribute to large‐scale acoustic monitoring of Orthoptera while recording bats. Further innovation of sound classifier algorithms is needed and would be aided by improved reference sound libraries from multiple locations spanning species’ ranges.

show abstract

Section: Introductionmentioning

confidence: 99%

Potential for coupling the monitoring of bush‐crickets with established large‐scale acoustic monitoring of bats

et al. 2017

View full text Add to dashboard Cite

show abstract

“…At the highest level, most research efforts advocate the extraction of sets of features from the data, and the use of these features as inputs for standard classification algorithms such as a decision tree, a Bayesian classifier or a neural network. As a concrete representative example, consider (Riede 2006), which introduces a system to recognize Orthoptera (the order of insects that includes grasshoppers, crickets, katydids (In British English, katydids are known as bush-crickets) and locusts). This method requires that we extract multiple features from the signal, including distancebetween-consecutive-pulses, pulse-length, frequency-contour-of-pulses, energy-contour-ofpulses, time-encoded-signal-of-pulses, etc.…”

Section: General Animal Sound Classificationmentioning

confidence: 99%

Monitoring and Mining Animal Sounds in Visual Space

2012

View full text Add to dashboard Cite

Monitoring animals by the sounds they produce is an important and challenging task, whether the application is outdoors in a natural habitat, or in the controlled environment of a laboratory setting.In the former case, the density and diversity of animal sounds can act as a measure of biodiversity. In the latter case, researchers often create control and treatment groups of animals, expose them to different interventions, and test for different outcomes. One possible manifestation of different outcomes may be changes in the bioacoustics of the animals.With such a plethora of important applications, there have been significant efforts to build bioacoustic classification tools. However, we argue that most current tools are severely limited. They often require the careful tuning of many parameters (and thus huge amounts of training data), are either too computationally expensive for deployment in resource-limited sensors, specialized for a very small group of species, or are simply not accurate enough to be useful.In this work we introduce a novel bioacoustic recognition/classification framework that mitigates or solves all of the above problems. We propose to classify animal sounds in the visual space, by treating the texture of their sonograms as an acoustic fingerprint using a recently introduced parameter-free texture measure as a distance measure. We further show that by searching for the most representative acoustic fingerprint, we can significantly outperform other techniques in terms of speed and accuracy.

show abstract

“…At the highest level, most research efforts advocate extracting sets of features from the data, and using these features as inputs to standard classification algorithms such as a decision tree, a Bayesian classifier or a neural network. As a concrete representative example, consider [24], which introduces a system to recognize Orthoptera (the order of insects that includes grasshoppers, crickets, katydids 2 and locusts). This method requires that we extract multiple features from the signal, including distance-between-consecutivepulses, pulse-length, frequency-contour-of-pulses, energy-contour-of-pulses, time-encoded-signal-ofpulses, etc.…”

Section: General Animal Sound Classificationmentioning

confidence: 99%

Monitoring and Mining Insect Sounds in Visual Space

Hao

Campana

Keogh

2012

Proceedings of the 2012 SIAM International Conference on Data Mining

View full text Add to dashboard Cite

Monitoring animals by the sounds they produce is an important and challenging task, whether the application is outdoors in a natural habitat, or in the controlled environment of a laboratory setting. In the former case the density and diversity of animal sounds can act as a measure of biodiversity. In the latter case, researchers often create control and treatment groups of animals, expose them to different interventions, and test for different outcomes. One possible manifestation of different outcomes may be changes in the bioacoustics of the animals. With such a plethora of important applications, there have been significant efforts to build bioacoustic classification tools. However, we argue that most current tools are severely limited. They often require the careful tuning of many parameters (and thus huge amounts of training data), they are too computationally expensive for deployment in resource-limited sensors, they are specialized for a very small group of species, or they are simply not accurate enough to be useful. In this work we introduce a novel bioacoustic recognition/classification framework that mitigates or solves all of the above problems. We propose to classify animal sounds in the visual space, by treating the texture of their spectrograms as an acoustic fingerprint using a recently introduced parameter-free texture measure as a distance measure. We further show that by searching for the most representative acoustic fingerprint we can significantly outperform other techniques in terms of speed and accuracy.

show abstract

Automated annotation of Orthoptera songs: first results from analysing the DORSA sound repository

Abstract: BioOne Complete (complete.BioOne.org) is a full-text database of 200 subscribed and open-access titles in the biological, ecological, and environmental sciences published by nonprofit societies, associations, museums, institutions, and presses.

Cited by 14 publications

References 7 publications

Potential for coupling the monitoring of bush‐crickets with established large‐scale acoustic monitoring of bats

Potential for coupling the monitoring of bush‐crickets with established large‐scale acoustic monitoring of bats

Monitoring and Mining Animal Sounds in Visual Space

Monitoring and Mining Insect Sounds in Visual Space

Contact Info

Product

Resources

About