A Survey On Denoising Bioacoustic Signals: Comparing Aerial And Underwater Signal Processing Techniques

Abstract

Bioacoustic signal processing has emerged as a critical field in biological monitoring, species identification, and ecological assessment. However, the presence of noise poses significant challenges to the accurate analysis of these signals in both terrestrial and aquatic environments. This survey paper provides a comprehensive review of denoising techniques applied to bioacoustic signals across aerial and underwater domains. We systematically categorize and compare traditional signal processing methods, statistical approaches, and modern machine learning techniques. Our analysis reveals that while fundamental principles of signal processing remain consistent across domains, the unique acoustic properties and noise characteristics of air and water necessitate specialized approaches. We further identify key research gaps and propose future directions, including multimodal fusion, adaptive real-time processing, and standardized evaluation frameworks. This survey serves as a resource for researchers and practitioners working at the intersection of signal processing and bioacoustics in diverse environmental contexts.