Audio Classification of Bird Species: A Statistical Manifold Approach

Briggs, Forrest; Raich, Raviv; Fern, Xiaoli Z.

doi:10.1109/icdm.2009.65

Cited by 64 publications

(42 citation statements)

References 19 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…29,30 BoW has become well-established in the field of computer vision, where it is also referred to as bagof-features. Recently, it has been adapted for the classification of time series in audio and speech recognition, [31][32][33] electroencephalogram (EEG) and electrocardiogram (ECG) signals, 34 and time series similarity. 35 In these works, time series are treated as text documents and sections extracted from the time series as words.…”

Section: Bag-of-words Representationmentioning

confidence: 99%

Toward a functional near-infrared spectroscopy-based monitoring of pain assessment for nonverbal patients

Rojas

Huang

2017

J. Biomed. Opt

View full text Add to dashboard Cite

Abstract. Pain diagnosis for nonverbal patients represents a challenge in clinical settings. Neuroimaging methods, such as functional magnetic resonance imaging and functional near-infrared spectroscopy (fNIRS), have shown promising results to assess neuronal function in response to nociception and pain. Recent studies suggest that neuroimaging in conjunction with machine learning models can be used to predict different cognitive tasks. The aim of this study is to expand previous studies by exploring the classification of fNIRS signals (oxyhaemoglobin) according to temperature level (cold and hot) and corresponding pain intensity (low and high) using machine learning models. Toward this aim, we used the quantitative sensory testing to determine pain threshold and pain tolerance to cold and heat in 18 healthy subjects (three females), mean age AE standard deviation (31.9 AE 5.5). The classification model is based on the bag-of-words approach, a histogram representation used in document classification based on the frequencies of extracted words and adapted for time series; two learning algorithms were used separately, K-nearest neighbor (K-NN) and support vector machines (SVM). A comparison between two sets of fNIRS channels was also made in the classification task, all 24 channels and 8 channels from the somatosensory region defined as our region of interest (RoI). The results showed that K-NN obtained slightly better results (92.08%) than SVM (91.25%) using the 24 channels; however, the performance slightly dropped using only channels from the RoI with K-NN (91.53%) and SVM (90.83%). These results indicate potential applications of fNIRS in the development of a physiologically based diagnosis of human pain that would benefit vulnerable patients who cannot self-report pain. © The Authors. Published by SPIE under a Creative Commons Attribution 3.0 Unported License. Distribution or reproduction of this work in whole or in part requires full attribution of the original publication, including its DOI.

show abstract

Section: Bag-of-words Representationmentioning

confidence: 99%

Toward a functional near-infrared spectroscopy-based monitoring of pain assessment for nonverbal patients

Rojas

Huang

2017

J. Biomed. Opt

View full text Add to dashboard Cite

show abstract

“…A diferencia de la representación en el dominio de la frecuencia de Fourier, en los MFCCs la unidad de medida son los mels, en donde la escala de frecuencia se ubica de forma logarítmica, con lo cual se busca que la representación se acerque mejor a la percepción auditiva humana [34].…”

Section: Métodos De Representaciónunclassified

“…La clasificación de vocalizaciones de animales tradicionalmente se ha relegado a enfoques cualitativos, estos requieren de un gran consumo de tiempo y frecuentemente están sesgados por la subjetividad [21]. En esta revisión se usa el término detección en los problemas cuyo objetivo es localizar la ubicación temporal de la emisión de la vocalización de una especie objetivo [4], [22], [23], [34], [35], [50]. El término clasificación se usa para etiquetar un canto [17], [21], [24], [30], para ello se usa un algoritmo conocido como clasificador, cuya entrada corresponde a un conjunto de valores relacionados con atributos o características del objeto; el clasificador requiere entrenarse previamente con un conjunto de datos estadísticamente representativos.…”

Section: Métodos De Clasificaciónunclassified

“…Estos modelos son un método probabilístico que ha sido usado amplia y exitosamente en diferentes escenarios de reconocimiento automatizado tales como la identificación de habla y hablante, la clasificación de señales sísmicas y la clasificación de secuencias de genes. Los HMMs resultan muy apropiados para el análisis de señales bioacústicas debido a su capacidad de tratar datos secuenciales con alta variabilidad (e, incluso, de diferentes longitudes) [34], [8], [32]. Estos cuentan con una clara y elegante interpretación bayesiana, además, se han desarrollado algoritmos rápidos y eficientes para realizar el entrenamiento y la prueba de los modelos [53].…”

Section: Métodos De Clasificaciónunclassified

See 1 more Smart Citation

Reconocimiento automatizado de señales bioacústicas: Una revisión de métodos y aplicaciones

2013

View full text Add to dashboard Cite

ResumenDurante la última década se ha publicado una gran cantidad de estudios de investigación y aplicaciones sobre monitoreo bioacústico automatizado; no obstante, tales estudios están dispersos en la literatura de ingeniería y ciencias biológicas. En este artículo se presentan conceptos fundamentales sobre monitoreo bioacústico automatizado; se revisan, comparan y categorizan -en una taxonomía de técnicas de DSP/PR-las contribuciones de las investigaciones y las aplicaciones publicadas; se sugieren algunas direcciones para investigaciones futuras y se resaltan los retos y las oportunidades relacionados con la instalación de esta tecnología en Colombia.

show abstract

“…Several methods have proven successful in correctly labeling the species of single birds in low-noise environments [4,1,2]. We propose a method of pre-processing and segmenting noisy field recordings of bird song, to isolate each bird syllable from the rest of the signal.…”

Section: Introductionmentioning

confidence: 99%

Time-frequency segmentation of bird song in noisy acoustic environments

Neal

Briggs

Raich

et al. 2011

2011 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP)

Self Cite

View full text Add to dashboard Cite

Recent work in machine learning considers the problem of identifying bird species from an audio recording. Most methods require segmentation to isolate each syllable of bird call in input audio. Energy-based time-domain segmentation has been successfully applied to low-noise, single-bird recordings. However, audio from automated field recorders contains too much noise for such methods, so a more robust segmentation method is required. We propose a supervised timefrequency audio segmentation method using a Random Forest classifier, to extract syllables of bird call from a noisy signal. When applied to a test data set of 625 field-collected audio segments, our method isolates 93.6% of the acoustic energy of bird song with a false positive rate of 8.6%, outperforming energy thresholding.

show abstract

Audio Classification of Bird Species: A Statistical Manifold Approach

Cited by 64 publications

References 19 publications

Toward a functional near-infrared spectroscopy-based monitoring of pain assessment for nonverbal patients

Toward a functional near-infrared spectroscopy-based monitoring of pain assessment for nonverbal patients

Reconocimiento automatizado de señales bioacústicas: Una revisión de métodos y aplicaciones

Time-frequency segmentation of bird song in noisy acoustic environments

Contact Info

Product

Resources

About