Bird species recognition from field recordings using HMM-based modelling of frequency tracks

Abstract

This paper presents an automatic system for detection of bird species in field recordings. A sinusoidal detection algorithm is employed to segment the acoustic scene into isolated spectro-temporal segments. Each segment is represented as a temporal sequence of frequencies of the detected sinusoid, referred to as frequency track. Each bird species is represented by a set of hidden Markov models (HMMs), each HMM modelling an individual type of bird vocalisation element. These HMMs are obtained in an unsupervised manner. The detection is based on a likelihood ratio of the test utterance against the target bird species and non-target background model. We explore on selection of cohort for modelling the background model, z-norm and t-norm score normalisation techniques and score compensation to deal with outlier data. Experiments are performed using over 40 hours of audio field recordings from 48 bird species plus an additional 16 hours of field recordings as impostor trials. Evaluations are performed using detection error trade-off plots. The equal error rate of 5% is achieved when impostor trials are non-target bird species vocalisations and 1.2% when using field recordings which do not contain bird vocalisations.

Automatic detection of bird species from audio field recordings using HMM-based modelling of frequency tracks

“…It starts with a brief review of the approach we employed for segmentation of the audio signal and extraction of frequency track features and then follows with unsupervised HMM-based modelling of individual vocalisation elements of each bird species. These two components of the system were introduced in our recent publications [4], [8], [10] where we refer the reader to for further details. We then describe the methods we employed for detection of bird species, including score normalisation techniques.…”

“…This representation was employed in a few earlier studies [1], [6] and also in our recent works [3], [4], [7], [8], [9], [10]. Among the acoustic modelling approaches, the most commonly used are Gaussian mixture models (GMM) [1], [3], hidden Markov models (HMMs) [1], [4], [6], [11], and decision trees [12]. Several studies focused on detection of specific bird species [13], [14], [15].…”

“…Typically, the first stage of an automatic system is to parse the acoustic signal into isolated spectro-temporal segments. This is often performed using an energy-based thresholding that requires an estimate of noise level, e.g., [1], or by decomposition into sinusoidal components [1], [2], [3], [4]. A variety of approaches to feature representation of the spectro-temporal segments and their modelling were explored.…”

“…In a case of tonal bird vocalisations, the use of a sinusoidal detection for segmentation also offers a natural way of representing the segment as a temporal sequence of the frequencies of the detected sinusoid, which we refer to as frequency track. This representation was employed in a few earlier studies [1], [6] and also in our recent works [3], [4], [7], [8], [9], [10]. Among the acoustic modelling approaches, the most commonly used are Gaussian mixture models (GMM) [1], [3], hidden Markov models (HMMs) [1], [4], [6], [11], and decision trees [12].…”

“…We use the sinusoidal detection algorithm introduced in [16] to parse the acoustic signal into isolated time-frequency segments and represent each segment using a temporal sequence of the frequency track features [4]. We employ HMM-based modelling of individual vocalisation elements, which are trained in an unsupervised manner.…”