Automated bird acoustic event detection and robust species classification

N on - invasive b ioacoustic monitoring is becoming increasingly popular for b iodiversity conservation. Two automated methods for acoustic classification of bird species currently used are frame - based methods, a model that uses Hidden Markov Models (HMMs), and event - based methods, a model consisting of descriptive measurements or restricted to tonal or harmonic vocalizations. In this work, we propose a new method for automated field recording anal ysis with improved automated segm entation and robust bird species classification . We used a Gaussian Mixture Model ( GMM ) - based frame selection with an event - energy - based sifting procedure that selected representative acoustic events . We employed a Mel , band - pass filter bank on each event ’ s spectrogram. T he output in each subband was parameterized by an autoregressive (AR) model , which result ed in a feature consisting of all model coefficients. Finally, a support vector machine (SVM) algorithm was used for classification . The significance o f the proposed method lies in the parameterized feature s depict ing the species - specific spectral pattern . This experiment used a control audio dataset and real - world audio dataset comprised of field recordings of eleven bird species from the X eno - canto A rc hive , consisting of 2762 bird acoustic events with 339 detected “ unknown ” events (corresponding to noise or unknown species vocalization s) . Compared with other recent approach es , our proposed method provide s comparable identification performance with respe ct to the eleven species of interest . Meanwhile, superior robu stness in real - world scenarios is achieved, which is expressed as the considerable improvement from 0.632 to 0.928 for the F - score metric regarding the “ unknown ” events . The advantage makes the proposed method more suitable for automated field recording analysis.