Sparse coding for scaled bioacoustics: From Humpback whale songs evolution to forest soundscape analyses

Glotin, Hervé; Sueur, Jérôme; Artières, Thierry; Adam, Olivier; Razik, Joseph

doi:10.1121/1.4805502

Cited by 5 publications

(8 citation statements)

References 0 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…We gathered four datasets, each representing a large amount of audio data and a large number of species to classify ( Table 1 ). Two of the datasets ( nips4b and lifeclef2014 ) consist of the publicly-released training data from bird classification challenges organised by the SABIOD project ( Glotin et al, 2013 ; Goëau et al, 2014 ). The nips4b dataset is multilabel (median 1 species per recording, range 0–6); the lifeclef2014 dataset is single-label but much larger.…”

Section: Methodsmentioning

confidence: 99%

“…One recent research project (named “SABIOD”) has provided a valuable stimulus to the research community by conducting classification challenges evaluated on large datasets of bird sounds collected in the wild, and with large numbers of species to recognise ( Glotin et al, 2013 ; Fodor, 2013 ; Goëau et al, 2014 ). The research reported in this paper benefits from the datasets made available through that project, as well as other datasets, to evaluate bird sound classification suitable for large-scale practical deployments.…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Automatic large-scale classification of bird sounds is strongly improved by unsupervised feature learning

Stowell

Plumbley

2014

PeerJ

225

181

View full text Add to dashboard Cite

Automatic species classification of birds from their sound is a computational tool of increasing importance in ecology, conservation monitoring and vocal communication studies. To make classification useful in practice, it is crucial to improve its accuracy while ensuring that it can run at big data scales. Many approaches use acoustic measures based on spectrogram-type data, such as the Mel-frequency cepstral coefficient (MFCC) features which represent a manually-designed summary of spectral information. However, recent work in machine learning has demonstrated that features learnt automatically from data can often outperform manually-designed feature transforms. Feature learning can be performed at large scale and “unsupervised”, meaning it requires no manual data labelling, yet it can improve performance on “supervised” tasks such as classification. In this work we introduce a technique for feature learning from large volumes of bird sound recordings, inspired by techniques that have proven useful in other domains. We experimentally compare twelve different feature representations derived from the Mel spectrum (of which six use this technique), using four large and diverse databases of bird vocalisations, classified using a random forest classifier. We demonstrate that in our classification tasks, MFCCs can often lead to worse performance than the raw Mel spectral data from which they are derived. Conversely, we demonstrate that unsupervised feature learning provides a substantial boost over MFCCs and Mel spectra without adding computational complexity after the model has been trained. The boost is particularly notable for single-label classification tasks at large scale. The spectro-temporal activations learned through our procedure resemble spectro-temporal receptive fields calculated from avian primary auditory forebrain. However, for one of our datasets, which contains substantial audio data but few annotations, increased performance is not discernible. We study the interaction between dataset characteristics and choice of feature representation through further empirical analysis.

show abstract

Section: Methodsmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

Automatic large-scale classification of bird sounds is strongly improved by unsupervised feature learning

Stowell

Plumbley

2014

PeerJ

225

181

View full text Add to dashboard Cite

show abstract

“…Sparse decomposition using dictionaries of atoms based on biologically informed time-frequency atoms such as Gabor and Gammatone functions–which are seen to resemble characteristics of cochlea filters–are intuitively attractive as they can provide a feature set which is oriented in a two dimensional time-frequency space with which to approximate the original signal. This has been shown to be more efficient than Fourier or wavelet representations ( Smith & Lewicki , 2005 ) and to provide effective and efficient input features in a range of audio discrimination tasks in everyday sounds ( Adiloglu et al , 2012 ), drum samples ( Scholler & Purwins , 2011 ) and similarity matching of bioacoustic data ( Glotin et al , 2013 ).…”

Section: Introductionmentioning

confidence: 99%

A new method for ecoacoustics? Toward the extraction and evaluation of ecologically-meaningful soundscape components using sparse coding methods

Eldridge

Casey

Moscoso

et al. 2016

PeerJ

View full text Add to dashboard Cite

Passive acoustic monitoring is emerging as a promising non-invasive proxy for ecological complexity with potential as a tool for remote assessment and monitoring (Sueur & Farina, 2015). Rather than attempting to recognise species-specific calls, either manually or automatically, there is a growing interest in evaluating the global acoustic environment. Positioned within the conceptual framework of ecoacoustics, a growing number of indices have been proposed which aim to capture community-level dynamics by (e.g., Pieretti, Farina & Morri, 2011; Farina, 2014; Sueur et al., 2008b) by providing statistical summaries of the frequency or time domain signal. Although promising, the ecological relevance and efficacy as a monitoring tool of these indices is still unclear. In this paper we suggest that by virtue of operating in the time or frequency domain, existing indices are limited in their ability to access key structural information in the spectro-temporal domain. Alternative methods in which time-frequency dynamics are preserved are considered. Sparse-coding and source separation algorithms (specifically, shift-invariant probabilistic latent component analysis in 2D) are proposed as a means to access and summarise time-frequency dynamics which may be more ecologically-meaningful.

show abstract

“…Mais ce regroupement est dépendant de l'initialisation, problème connu des algorithmes KNN. Dans l'article de (Glotin et al, 2013), il est proposé de regrouper des unités de chant extraites de décomposition parcimonieuses pour obtenir des représentations plus efficaces en suivi d'animaux, c'est ce que nous approfondissons dans ce papier pour le cas des chants de baleine à bosses. Cet article propose pour la première fois un codage parcimonieux entièrement automatique du chant afin de déterminer leurs composantes stables par rapport à celles qui évoluent, à des échelles de temps différentes.…”

Section: Introductionunclassified

Décomposition parcimonieuse des chants de cétacés pour leur suivi

Doh¹,

Razik²,

Paris³

et al. 2013

Traitement du signal

Self Cite

View full text Add to dashboard Cite

les baleines à bosse mâles émettent des vocalises organisées et, pour certaines, répétées formant ainsi le leitmotiv d'un chant. Principalement, dans le but de mieux appréhender le comportement de ces baleines et notamment les interactions entre individus (mâle/mâle, mâle/femelle), plusieurs études sont actuellement menées sur ces chants. Dans cette étude, nous nous intéressons aux unités sonores, vocalises séparées par 2 silences, qui composent ces chants, à leurs récurrences, et à leurs structurations. Cependant, tous ces paramètres dépendent de l'année et du lieu d'enregistrement. Des travaux antérieurs ont souligné la nécessité de méthodes objectives pour la classification de ces unités sonores. L'analyse détaillée des vocalisations a montré que les caractéristiques d'une unité peuvent changer brusquement pendant toute sa durée, ce qui les rend difficiles à caractériser et à grouper systématiquement. Cet article propose un codage parcimonieux des chants afin de déterminer leurs composantes stables de celles qui varient, pour différentes échelles de temps. Une définition de la complexité du code est également proposée afin de séparer les composantes Traitement du signal-n o 3-4-5/2013, 219-242

show abstract

Sparse coding for scaled bioacoustics: From Humpback whale songs evolution to forest soundscape analyses

Cited by 5 publications

References 0 publications

Automatic large-scale classification of bird sounds is strongly improved by unsupervised feature learning

Automatic large-scale classification of bird sounds is strongly improved by unsupervised feature learning

A new method for ecoacoustics? Toward the extraction and evaluation of ecologically-meaningful soundscape components using sparse coding methods

Décomposition parcimonieuse des chants de cétacés pour leur suivi

Contact Info

Product

Resources

About