Unsupervised Dolphin Whistle Signal Dereverberation Based on β-Divergence Convolutive Nonnegative Matrix Factorization With Mixed Constraints

Reverberation significantly affects the feature extraction and detection performance of dolphin whistle signals in underwater passive acoustic monitoring (PAM). To mitigate this interference, we propose a unsupervised signal deconvolution method based on generalized $\beta $ -divergence convolutive nonnegative matrix factorization with mixed constraints ( $\beta $ -CNMF-MC) for dolphin whistles. First, we employ dictionary learning to capture the contour structures of whistle spectrograms from t–f segments of observed dolphin calls. Then, using the learned dictionary, we reconstruct the pure whistle signals by applying mixed constraints of underwater acoustic channel (UAC) characteristics. Both stages are formulated as the minimization of a unified cost function incorporating a mixed penalty term, solved via an agent-assisted optimization algorithm tailored for local minima convergence. Among them, the $\beta $ -divergence is used as a nonnegative fidelity term, utilizing the inherent sparsity to enhance the fidelity of the time-frequency (t–f) representation of the whistle signals amidst reverberation. The optimal parameter selection of the proposed method is identified using a two-stage optimization algorithm to ensure the most accurate dereverberation outcome. The proposed method was validated on a dataset involving two Southern Bottlenose Dolphins, with experimental results indicating robust performance across varying degrees of reverberant interference.

I. Introduction

Underwater passive acoustic monitoring (PAM) has become a popular method for studying the behavior and distribution of marine mammals, especially dolphins, due to its continuous monitoring capability and high adaptability [1], [2]. The complex and highly developed communication systems of dolphins reflect the complexity of their social relationships [3]. The communication signals emitted by dolphins have highly variable amplitude-modulated (AM) pitch and frequency-modulated (FM) pitch, also known as whistle signals [4]. Whistle signals exhibit a highly structured and sparse representation because of the physical limitations of the dolphin vocal apparatus [5]. Meanwhile, whistles vary significantly between and within populations, and specific categories can be used to convey specific information about the identity of individual species [6]. Meanwhile, whistle signals of different amplitude-frequency modulation (AM-FM) modes correspond to different behavioral information of dolphins [7]. As such, PAM technology provides valuable insights into dolphin habitat locations, social interactions, and population densities while minimizing the negative impacts of human activities [8], [9]. Given these advantages, PAM has become a critical tool for marine biodiversity and ecological conservation efforts.