How to Correctly Evaluate an Automatic Bioacoustics Classification Method

Colonna, Juan Gabriel; Gama, João; Nakamura, Eduardo F.

doi:10.1007/978-3-319-44636-3_4

Cited by 9 publications

(8 citation statements)

References 19 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…Each point corresponds to one sample from the mosquito wingbeat dataset in the plots, and the different colors denote the different classes. We employed 20 MFCCs and 12 LPCs to represent the mosquito wingbeat audio samples since these parameters are commonly used in literature [13], [33]. According to the t-SNE plots, LPC clusters are visually more compact but exhibiting many outliers; differently, the MFCC clusters appear more separable than LPC.…”

Section: Separability Of Mssa Dssc and Related Methodsmentioning

confidence: 99%

See 1 more Smart Citation

Discriminative Singular Spectrum Analysis for Bioacoustic Classification

Gatto¹,

Santos²,

Colonna³

et al. 2020

Interspeech 2020

View full text Add to dashboard Cite

Automatic analysis of bioacoustic signals is a fundamental tool to evaluate the vitality of our planet. Frogs and bees, for instance, may act like biological sensors providing information about environmental changes. This task is fundamental for ecological monitoring and includes many challenges such as nonuniform signal length processing, degraded target signal due to environmental noise, and the scarcity of the labeled samples for training machine learning. To tackle these challenges, we present a bioacoustic signal classifier equipped with a discriminative mechanism to extract useful features for analysis and classification efficiently. The proposed classifier does not require a large amount of training data and handles nonuniform signal length natively. Unlike current bioacoustic recognition methods, which are task-oriented, the proposed model relies on transforming the input signals into vector subspaces generated by applying Singular Spectrum Analysis (SSA). Then, a subspace is designed to expose discriminative features. The proposed model shares end-to-end capabilities, which is desirable in modern machine learning systems. This formulation provides a segmentation-free and noise-tolerant approach to represent and classify bioacoustic signals and a highly compact signal descriptor inherited from SSA. The validity of the proposed method is verified using three challenging bioacoustic datasets containing anuran, bee, and mosquito species. Experimental results on three bioacoustic datasets have shown the competitive performance of the proposed method compared to commonly employed methods for bioacoustics signal classification in terms of accuracy.

show abstract

Section: Separability Of Mssa Dssc and Related Methodsmentioning

confidence: 99%

“…The anuran dataset [33] consists of 60 recordings of 10 different species of frogs with varying record lengths collected under noise conditions. The number of records per species ranges from 3 to 11.…”

Section: A Datasetsmentioning

confidence: 99%

Discriminative Singular Spectrum Analysis for Bioacoustic Classification

Gatto¹,

Santos²,

Colonna³

et al. 2020

Interspeech 2020

View full text Add to dashboard Cite

show abstract

“…Mel-frequency cepstral coefficients (MFCCs) are features in the bioacoustics domain [10,24,25]. First, the STFT is calculated based on the raw signal.…”

Section: Mel-frequency Cepstral Coefficients (Mfccs)mentioning

confidence: 99%

“…Accordingly, this dataset was suitable for testing the proposed pattern recognition system. As there have been several works [18,25] already done on event detection itself, for the evaluation of the proposed recognition system, we assumed that the identified events were available to be input into the system. For the experiments, bird experts listened to the audio files and recorded the different events' start times and end times.…”

Section: Data Sourcementioning

confidence: 99%

Identifying Patterns of Human and Bird Activities Using Bioacoustic Data

Garg

Brown

2019

Forests

View full text Add to dashboard Cite

In general, humans and animals often interact within the same environment at the same time. Human activities may disturb or affect some bird activities. Therefore, it is important to monitor and study the relationships between human and animal activities. This paper proposed a system able not only to automatically classify human and bird activities using bioacoustic data, but also to automatically summarize patterns of events over time. To perform automatic summarization of acoustic events, a frequency–duration graph (FDG) framework was proposed to summarize the patterns of human and bird activities. This system first performs data pre-processing work on raw bioacoustic data and then applies a support vector machine (SVM) model and a multi-layer perceptron (MLP) model to classify human and bird chirping activities before using the FDG framework to summarize results. The SVM model achieved 98% accuracy on average and the MLP model achieved 98% accuracy on average across several day-long recordings. Three case studies with real data show that the FDG framework correctly determined the patterns of human and bird activities over time and provided both statistical and graphical insight into the relationships between these two events.

show abstract

“…Primeira etapa da avaliação: O experimento foi realizado em duas etapas, para poder avaliar o erro de generalização esperado pelo modelo de classificação (como recomendado em [Colonna et al 2016a]). A primeira etapa utiliza validação cruzada por áudio, onde a cada iteração, todos os frames pertencentes ao mesmo áudio de motosserra neste caso, são separados da classe positiva e utilizados como conjunto de teste.…”

Section: Metodologia Experimentalunclassified

Sensor Acústico para Detecção de Desmatamento Ilegal na Floresta Amazônica

Seabra¹,

Colonna²,

Nakamura³

2017

Anais Do Simpósio Brasileiro De Computação Ubíqua E Pervasiva (SBCUP)

Self Cite

View full text Add to dashboard Cite

Neste trabalho, apresentamos um método para detecção de sons de motosserras, com o objetivo de auxiliar no combate de extração ilegal de madeira nas florestas tropicais. Em nossa abordagem utilizamos um método de classificação de uma classe para detectar apenas o som de interesse e rejeitar todos os outros sons, sejam estes naturais ou artificiais, tais como: chuva, vento, diversas vocalizações de animais, fala humana, e motores de barcos. O método de classificação escolhido é o Support Vector Data Description (SVDD), que consiste em criar uma hiperesfera ao redor dos pontos que representam a classe alvo no espaço de características acústicas. Aplicando a transformada Wavelet, para a extração das características acústicas, conseguimos processar o sinal de entrada em tempo real. Assim, o método proposto consegue processar e classificar osáudios em tempo real com baixa complexidade computacional, o que beneficia o uso de sensores por economizar processamento, reduzindo o consumo de bateria. Através dos experimentos realizados, concluímos que nossa abordagem é eficiente diferenciando sons de motosserras de sons naturais (AUC = 96%), diferenciando motosserras e sons artificiais de sons naturais (AUC = 95%), mas perde eficiência quando comparada com sons artificiais (AUC = 77%).

show abstract

How to Correctly Evaluate an Automatic Bioacoustics Classification Method

Cited by 9 publications

References 19 publications

Discriminative Singular Spectrum Analysis for Bioacoustic Classification

Discriminative Singular Spectrum Analysis for Bioacoustic Classification

Identifying Patterns of Human and Bird Activities Using Bioacoustic Data

Sensor Acústico para Detecção de Desmatamento Ilegal na Floresta Amazônica

Contact Info

Product

Resources

About