Automated bird acoustic event detection and robust species classification

Zhao, Zhao; Zhang, Saihua; Xu, Zhihai; Bellisario, Kristen; Nagamatsu, Dai; Omrani, Hichem; Pijanowski, Bryan C.

doi:10.1016/j.ecoinf.2017.04.003

Cited by 94 publications

(67 citation statements)

References 44 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“… BIRDZ, the control and real-world audio dataset used in [24]. The real-world recordings were  CAT, the cat sound dataset was presented in [28,50].…”

Section: Resultsmentioning

confidence: 99%

Data augmentation approaches for improving animal audio classification

Nanni

Maguolo

Paci

2020

Ecological Informatics

127

View full text Add to dashboard Cite

In this paper we present ensembles of classifiers for automated animal audio classification, exploiting different data augmentation techniques for training Convolutional Neural Networks (CNNs). The specific animal audio classification problems are i) birds and ii) cat sounds, whose datasets are freely available. We train five different CNNs on the original datasets and on their versions augmented by four augmentation protocols, working on the raw audio signals or their representations as spectrograms. We compared our best approaches with the state of the art, showing that we obtain the best recognition rate on the same datasets, without ad hoc parameter optimization.Our study shows that different CNNs can be trained for the purpose of animal audio classification and that their fusion works better than the stand-alone classifiers. To the best of our knowledge this is the largest study on data augmentation for CNNs in animal audio classification audio datasets using the same set of classifiers and parameters. Our MATLAB code is available at https://github.com/LorisNanni .

show abstract

“… BIRDZ, the control and real-world audio dataset used in [24]. The real-world recordings were  CAT, the cat sound dataset was presented in [28,50].…”

Section: Resultsmentioning

confidence: 99%

Data augmentation approaches for improving animal audio classification

Nanni

Maguolo

Paci

2020

Ecological Informatics

127

View full text Add to dashboard Cite

show abstract

“…Likewise, autonomous recording units have been used to detect the presence and spatial distribution of a variety of seabird species (Buxton and Jones 2012, Cragg et al 2015, Harvey et al 2016. In contrast to manual techniques for species identification, the development of automated machine learning methodologies (Bardeli et al 2010, Digby et al 2013, de Oliveira et al 2015, Stowell et al 2016, Zhao et al 2017 has the potential to rapidly advance the use of unstructured acoustic monitoring in ornithological research (Gorrepati et al 2012).…”

Section: Origins Of Big Data In Ornithologymentioning

confidence: 99%

Opportunities and challenges for big data ornithology

et al. 2018

View full text Add to dashboard Cite

Recent advancements in information technology and data acquisition have created both new research opportunities and new challenges for using big data in ornithology. We provide an overview of the past, present, and future of big data in ornithology, and explore the rewards and risks associated with their application. Structured data resources (e.g., North American Breeding Bird Survey) continue to play an important role in advancing our understanding of bird population ecology, and the recent advent of semistructured (e.g., eBird) and unstructured (e.g., weather surveillance radar) big data resources has promoted the development of new empirical perspectives that are generating novel insights. For example, big data have been used to study and model bird diversity and distributions across space and time, explore the patterns and determinants of broad-scale migration strategies, and examine the dynamics and mechanisms associated with geographic and phenological responses to global change. The application of big data also holds a number of challenges wherein high data volume and dimensionality can result in noise accumulation, spurious correlations, and incidental endogeneity. In total, big data resources continue to add empirical breadth and detail to ornithology, often at very broad spatial extents, but how the challenges underlying this approach can best be mitigated to maximize inferential quality and rigor needs to be carefully considered.

show abstract

“…Building such ensembles is motivated by two observations: 1) it is well known that ensembles of neural networks generally perform better than stand-alone models due to the instability of the training process [49], and 2) it has been shown in other classification tasks that an ensemble of multiple networks trained with different augmentation protocols performs much better than do stand-alone networks [50]. The scores of the neural networks trained here are combined by sum rule, and the proposed approach is tested across three different audio classification datasets: domestic cat sound classification ( [28]), bird call classification [51], and environmental classification [4]. Our experiments were designed both to compare and to maximize performance by varying sets of data augmentation methods with different image representations of the audio signals.…”

Section: Introductionmentioning

confidence: 99%

Ensemble of convolutional neural networks to improve animal audio classification

Nanni

Costa

Aguiar

et al. 2020

J AUDIO SPEECH MUSIC PROC.

View full text Add to dashboard Cite

Recently, deep learning classifiers have proven even more robust in pattern recognition and classification than have texture analysis techniques. With the broad availability of relatively inexpensive Graphics Processing Units (GPUs), many researchers have begun applying deep learning techniques to visual representations of acoustic traces. Preselected or handcrafted descriptors, such as LBP, are not necessary for deep learners since they learn salient features during the training phase. Deep learners, moreover, are uniquely suited to handling visual representations of audio because many of the most famous deep classifiers, such as Convolutional Neural Networks (CNN), require matrices as their input. Humphrey and Bello [17, 18] were among the first to apply CNNs to audio images for music classification and, as a result, succeeded in redefining the state of the art in automatic chord detection and recognition. In the same year, Nakashika et al. [19] reported converting spectrograms to GCLM maps to train CNNs to performed music genre classification on the GTZAN dataset [20]. Later, Costa et al. [21] fused a CNN with the traditional pattern recognition framework of training SVMs on LBP features to classify the LMD dataset. These works exceeded traditional classification results on these genre datasets. Up to this point, most work in audio classification has applied the latest advances in machine learning to the problem of sound classification and recognition without modifying the classification process to make it singularly suitable for sound recognition. An early exception to the generic approach is found in the work of Sigtia and Dixon [22], who adjusted CNN parameters and structures in such a way as to reduce the time it took to train a set of audio images. Time reduction was accomplished by replacing

show abstract

Automated bird acoustic event detection and robust species classification

Cited by 94 publications

References 44 publications

Data augmentation approaches for improving animal audio classification

Data augmentation approaches for improving animal audio classification

Opportunities and challenges for big data ornithology

Ensemble of convolutional neural networks to improve animal audio classification

Contact Info

Product

Resources

About