Compact representation of temporal processes in echosounder time series via matrix decomposition

Biodiversity assessment using passive acoustic recordings from off-reef location-Unsupervised learning to classify fish vocalization

We present the quantitative characterization of Grande Island's off-reef acoustic environment within the Zuari estuary during the pre-monsoon period. Passive acoustic recordings reveal prominent fish choruses. Detailed characteristics of the call employing oscillograms and individual fish call parameters of the segmented data include vocal groups such as Sciaenidae, Terapon theraps, and planktivorous as well as invertebrate sounds, e.g., snapping shrimp. We calculated biodiversity parameters (i) Acoustic Evenness Index (AEI), (ii) Acoustic Complexity Index (ACI), and mean sound pressure level (SPLrms) for three frequency bands such as full band (50-22 050 Hz), the low-frequency fish band (100-2000 Hz), and the high-frequency shrimp band (2000-20 000 Hz). Here, ACI and AEI metrics characterize the location's soundscape data effectively indicating increased biodiversity of fish species for both the low-frequency and high-frequency bands. Whereas variations for SPLrms are prominent for three frequency bands. Moreover, we employ unsupervised classification through a hybrid technique comprising principal component analysis (PCA) and K-means clustering for data features of four fish sound types. Employed PCA for dimensionality reduction and related K-means clustering successfully provides 96.20%, 76.81%, 100.00%, and 86.36% classification during the dominant fish chorus. Overall, classification performance (89.84%) is helpful in the real-time monitoring of the fish stocks in the ecosystem.

Introduction to the special issue on machine learning in acoustics

The use of machine learning (ML) in acoustics has received much attention in the last decade. ML is unique in that it can be applied to all areas of acoustics. ML has transformative potentials as it can extract statistically based new information about events observed in acoustic data. Acoustic data provide scientific and engineering insight ranging from biology and communications to ocean and Earth science. This special issue included 61 papers, illustrating the very diverse applications of ML in acoustics.

Deep embedded clustering of coral reef bioacoustics

Deep clustering was applied to unlabeled, automatically detected signals in a coral reef soundscape to distinguish fish pulse calls from segments of whale song. Deep embedded clustering (DEC) learned latent features and formed classification clusters using fixed-length power spectrograms of the signals. Handpicked spectral and temporal features were also extracted and clustered with Gaussian mixture models (GMM) and conventional clustering. DEC, GMM, and conventional clustering were tested on simulated datasets of fish pulse calls (fish) and whale song units (whale) with randomized bandwidth, duration, and SNR. Both GMM and DEC achieved high accuracy and identified clusters with fish, whale, and overlapping fish and whale signals. Conventional clustering methods had low accuracy in scenarios with unequal-sized clusters or overlapping signals. Fish and whale signals recorded near Hawaii in February-March 2020 were clustered with DEC, GMM, and conventional clustering. DEC features demonstrated the highest accuracy of 77.5% on a small, manually labeled dataset for classifying signals into fish and whale clusters.