Predicting species identity of bumblebees through analysis of flight buzzing sounds

Gradišek, Anton; Slapničar, Gašper; Šorn, Jure; Luštrek, Mitja; Gams, Matjaž; Grad, Janez

doi:10.1080/09524622.2016.1190946

Cited by 40 publications

(42 citation statements)

References 19 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…However, in terms of species identification, bee species were classified with relatively low accuracy (0.7-0.9 in precision and recall), although the hornet species (V. s. xanthoptera ) could be accurately classified (1.00 in precision and recall). Regarding bee species discrimination, Gradišek et al (2017) tried to identify 12 species of bumblebees using acoustic analysis, and found that the accuracy of identification varied between species (0.0-1.00 in precision and recall) (Caliculated from Table II in Gradišek et al 2017). In their study, a few species (such as brown-banded carder bee, B. humilis , queens or early bumble bee, B. pratorum, workers) were more accurately identified (precision and recall both > 0.9), and most of the species were identified with precision and recall between 0.50-0.85 in their validation of the model using the training dataset (Caliculated from Table II in Gradišek et al 2017).…”

Section: Discussionmentioning

confidence: 99%

“…There were clear differences in the frequency spectra and the harmonic components between their flight sounds and the background sounds ( Figure 1). Therefore, we used mel-frequency cepstral coefficients (MFCC) to describe the acoustic characteristic feature values of the different types of sounds, because MFCC was one of the most frequently used feature values in identifying sounds from different insects in previous studies, such as Orthoptera (Chaves et al 2012;Zhang et al 2012), Cicadae (Zilli et al 2014), and some bumble bees (Gradišek et al 2017). Basically, MFCC describes the timbre of sounds and is calculated using the following steps: (1) slicing the original sound into frames, (2) applying a window function to each frame, (3) applying Fourier transformation to each frame and obtaining the power spectrum of each frame, (4) applying mel-scale filter banks to the frames, and (5) applying a discrete cosine transformation (DCT).…”

Section: Methodsmentioning

confidence: 99%

“…Towsey et al (2014) demonstrated that the use of acoustic indices could identify the cicada chorus in the natural environment, and Lampson et al (2013) developed automatic identification methods for stink bugs (Euschistus servus and Nezara viridula ) using acoustic analysis of intraspecific substrateborne vibrational signals. Recently, Gradišek et al (2017) tried to discriminate bumblebee species using the acoustic features of their flight sounds, and found that the different species differed in their flight sounds. In this way, acoustic/ vibrational-based monitoring technology is becoming popular, but previous studies have focused on the Cicadae and Orthoptera (Obrist et al 2010) or specific bee species such as bumble bees (De Luca et al 2014;Gradišek et al 2017;Miller-Struttmann et al 2017), and, to our knowledge, there are still few studies that focus on identifying different types of bees by their sounds.…”

Section: Introductionmentioning

confidence: 99%

“…Recently, Gradišek et al (2017) tried to discriminate bumblebee species using the acoustic features of their flight sounds, and found that the different species differed in their flight sounds. In this way, acoustic/ vibrational-based monitoring technology is becoming popular, but previous studies have focused on the Cicadae and Orthoptera (Obrist et al 2010) or specific bee species such as bumble bees (De Luca et al 2014;Gradišek et al 2017;Miller-Struttmann et al 2017), and, to our knowledge, there are still few studies that focus on identifying different types of bees by their sounds. Especially, in practical sense, distinguishing predators and pollinators are important for beekeepers or ecologist so that investigating whether the acoustic analysis can identify bee species into functional group is informative.…”

Section: Introductionmentioning

confidence: 99%

See 3 more Smart Citations

Automated classification of bees and hornet using acoustic analysis of their flight sounds

Kawakita

Ichikawa

2019

Apidologie

View full text Add to dashboard Cite

To investigate how to accurately identify bee species using their sounds, we conducted acoustic analysis to identify three pollinating bee species (Apis mellifera , Bombus ardens , Tetralonia nipponensis ) and a hornet (Vespa simillima xanthoptera ) by their flight sounds. Sounds of the insects and their environment (background noises and birdsong) were recorded in the field. The use of fundamental frequency and mel-frequency cepstral coefficients to describe feature values of the sounds, and supported vector machines to classify the sounds, correctly distinguished sound samples from environmental sounds with high recalls and precision (0.96-1.00). At the species level, our approach could classify the insect species with relatively high recalls and precisions (0.7-1.0). The flight sounds of V.s. xanthoptera , in particular, were perfectly identified (precision and recall 1.0). Our results suggest that insect flight sounds are potentially useful for detecting bees and quantifying their activity. species classification / Hymenoptera / machine learning / acoustic analysis

show abstract

Section: Discussionmentioning

confidence: 99%

Section: Methodsmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

See 2 more Smart Citations

Automated classification of bees and hornet using acoustic analysis of their flight sounds

Kawakita

Ichikawa

2019

Apidologie

View full text Add to dashboard Cite

show abstract

“…43,44 Ensemble learning is a well-known method of combining classifiers to improve results by reducing the variance of the classification decision through the use of multiple classifiers that learn from different data, with wellknown examples such as random forest 201 and adaptive boosting. 202 These techniques have been leveraged by the bioacoustics community, such as the work by Gradiek et al (2016) 203 that used random forests to distinguish bumble bee species based on characteristics of their buzz.…”

Section: Bioacousticsmentioning

confidence: 99%

Machine learning in acoustics: Theory and applications

Bianco

Gerstoft

Traer

et al. 2019

The Journal of the Acoustical Society of America

392

126

View full text Add to dashboard Cite

Acoustic data provide scientific and engineering insights in fields ranging from biology and communications to ocean and Earth science. We survey the recent advances and transformative potential of machine learning (ML), including deep learning, in the field of acoustics. ML is a broad family of statistical techniques for automatically detecting and utilizing patterns in data. Relative to conventional acoustics and signal processing, ML is data-driven. Given sufficient training data, ML can discover complex relationships between features and desired labels or actions, or between features themselves. With large volumes of training data, ML can discover models describing complex acoustic phenomena such as human speech and reverberation. ML in acoustics is rapidly developing with compelling results and significant future promise. We first introduce ML, then highlight ML developments in five acoustics research areas: source localization in speech processing, source localization in ocean acoustics, bioacoustics, seismic exploration, and environmental sounds in everyday scenes.

show abstract

Independent origins of a novel atympanic middle ear system within Chamaeleonidae

Olroyd

2022

The Anatomical Record

View full text Add to dashboard Cite

The evolution of the vertebrate ear is a complicated story of convergence, cooption, loss of function, and occasional regaining of said function. An incredible variety of structures has been adopted as sound receptors, but only chameleons are known to have a bony airborne sound receiver. In some chameleons, the pterygoid bone captures sound vibrations and relays them to the inner ear via a connection to the extracolumella. The distribution of this unique hearing system has not been examined across Chamaeleonidae. Here, I report on dissections on 12 species across four genera and describe their middle ear anatomy for the first time. Half of these species were found to have a link between their extracolumella and pterygoid, and ancestral state reconstruction supports four independent acquisitions of this novel sound-conduction pathway. Species with this pathway tend to have a gular pouch, which seems to produce biotremors and possibly airborne sound, suggesting that this hearing system plays some role in intraspecific communication. Three species were also μ-CT scanned using enhanced contrast to investigate differences in the musculature surrounding the middle ear cavity. In species with a middle ear connected to the pterygoid, the muscles directly lateral to the pterygoid insert farther anterior onto the mandible, which may serve to minimize dampening of vibrations on the pterygoid. Together, these data suggest that the ear plays a more significant role in the lives of some chameleons than has been recognized, and that parallelism is common in the evolution of the ear.

show abstract

Predicting species identity of bumblebees through analysis of flight buzzing sounds

Cited by 40 publications

References 19 publications

Automated classification of bees and hornet using acoustic analysis of their flight sounds

Automated classification of bees and hornet using acoustic analysis of their flight sounds

Machine learning in acoustics: Theory and applications

Independent origins of a novel atympanic middle ear system within Chamaeleonidae

Contact Info

Product

Resources

About